instruction
stringclasses 1
value | output
stringlengths 67
772
| input
stringlengths 28
631
|
---|---|---|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the River Sava
#E1 = Wikipedia[River Sava]
Plan: Find out the name of the capital city on the River Sava
#E2 = LLM[What is the name of the capital city on the River Sava? Given context: #E1] | Which European capital city lies on the River Sava? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the Godfather trilogy
#E1 = Wikipedia[Godfather trilogy]
Plan: Find out who directed the trilogy
#E2 = LLM[Who directed the Godfather trilogy? Given context: #E1] | Who directed the Godfather trilogy of films? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about archipelago
#E1 = Wikipedia[Archipelago] | An archipelago is a group of what? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the history of the Olympic Games
#E1 = Wikipedia[Olympic Games]
Plan: Search for more information about the history of television coverage of the Olympic Games
#E2 = Wikipedia[Television coverage of the Olympic Games]
Plan: Find out in which year live television coverage of the Olympic Games was available internationally for the first time.
#E3 = LLM[In which year was live television coverage of the Olympic Games available internationally for the first time? Given context: #E1 and #E2] | In which year was live television coverage of the Olympic Games available internationally for the first time? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Research the date of the 1st
#E1 = Wikipedia[1st of month]
Plan: Find out the day of the week of the 1st
#E2 = LLM[What day of the week was the 1st? Given context: #E1] | Imagine it's Friday the 13th, what day of the week was the 1st? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about J. Rodman Williams.
#E1 = Wikipedia[J. Rodman Williams]
Plan: Find out which university J. Rodman Williams was a professor at.
#E2 = LLM[Which university was J. Rodman Williams a professor at? Given context: #E1]
Plan: Search for more information about the university.
#E3 = Wikipedia[#E2]
Plan: Find out when the university was founded.
#E4 = LLM[When was #E2 founded? Given context: #E3] | When was the University founded where J. Rodman Williams was a professor ?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | in the movie Clueless?
Plan: Search for more information about the actress.
#E1 = Wikipedia[Alicia Silverstone]
Plan: Find out the stage debut of the actress.
#E2 = LLM[What was stage debut for Alicia Silverstone? Given context: #E1] | What was stage debut for the actress that voiced her role in the Happy Feet game and played Tai Frasier
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about G8 Conferences
#E1 = Wikipedia[G8 Conferences]
Plan: Search for more information about the 2009 G8 Conference
#E2 = Wikipedia[2009 G8 Conference]
Plan: Find out the name of the Italian town where the 2009 G8 Conference was held.
#E3 = LLM[What is the name of the Italian town where the 2009 G8 Conference was held? Given context: #E1 and #E2] | In which Italian town was the World G8 Conference held in 2009? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the 1975 Wimbledon Men's Singles Final
#E1 = Wikipedia[1975 Wimbledon Men's Singles Final]
Plan: Find out who beat Jimmy Connors in the 1975 Wimbledon Men's Singles Final
#E2 = LLM[Who beat Jimmy Connors in the 1975 Wimbledon Men's Singles Final? Given context: #E1] | Who unexpectedly beat Jimmy Connors in the 1975 Wimbledon Men's Singles Final? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Martin Fry and Mike Shinoda.
#E1 = Wikipedia[Martin Fry]
#E2 = Wikipedia[Mike Shinoda]
Plan: Compare the scope of profession of Martin Fry and Mike Shinoda.
#E3 = LLM[Who has more scope of profession, Martin Fry or Mike Shinoda? Given context: #E1 and #E2] | Who has more scope of profession, Martin Fry or Mike Shinoda?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about mobile software products.
#E1 = Wikipedia[mobile software products]
Plan: Find out the company that develops mobile software products.
#E2 = LLM[What company develops mobile software products in the discipline that focuses upon the efficient and effective delivery of support services? Given context: #E1] | What company develops mobile software products in the discipline that focuses upon the efficient and effective delivery of support services?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Manet
#E1 = Wikipedia[Manet]
Plan: Read the Wikipedia article to find out his Christian names
#E2 = LLM[Read the Wikipedia article to find out Manet's Christian names. Given context: #E1] | By which of his Christian names is the French artist Manet (1832-1883) known? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Whitehaven Mansions
#E1 = Wikipedia[Whitehaven Mansions]
Plan: Search for more information about Sandhurst Sq
#E2 = Wikipedia[Sandhurst Sq]
Plan: Find out the address of Apt 56B
#E3 = LLM[What is the address of Apt 56B? Given context: #E1 and #E2] | Apt 56B, Whitehaven Mansions, Sandhurst Sq, London |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Elvis Presley
#E1 = Wikipedia[Elvis Presley]
Plan: Search for more information about the Memphis Mafia
#E2 = Wikipedia[Memphis Mafia] | Whose bodyguards were known as the Memphis Mafia |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Atomic Ed and the Black Hole.
#E1 = Wikipedia[Atomic Ed and the Black Hole]
Plan: Find out the release date of Atomic Ed and the Black Hole.
#E2 = LLM[When was Atomic Ed and the Black Hole released? Given context: #E1]
Plan: Search for more information about Flying Padre.
#E3 = Wikipedia[Flying Padre]
Plan: Find out the release date of Flying Padre.
#E4 = LLM[When was Flying Padre released? Given context: #E3] | Which film was released first, Atomic Ed and the Black Hole or Flying Padre?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Smallpox
#E1 = Wikipedia[Smallpox]
Plan: Search for more information about the history of Smallpox vaccine
#E2 = Wikipedia[Smallpox vaccine]
Plan: Find out the name of the developer of the first successful Smallpox vaccine.
#E3 = LLM[Who developed the first successful Smallpox vaccine? Given context: #E1 and #E2] | Who developed the first suvvessful vaccine for Smallpox ? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Niel Hamilton.
#E1 = Wikipedia[Niel Hamilton]
Plan: Find out the name of Niel Hamilton's wife.
#E2 = LLM[What is the name of Niel Hamilton's wife? Given context: #E1]
Plan: Search for more information about the cooking show that inspired Posh Nosh.
#E3 = Wikipedia[Posh Nosh]
Plan: Find out who pitched the pilot for the cooking show that inspired Posh Nosh.
#E4 = LLM[Who pitched the pilot for the cooking show that inspired Posh Nosh? Given context: #E3] | Who was the wife of Niel Hamilton who also pitched the pilot to a cooking show that inspired Posh Nosh? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about geometrical figures
#E1 = Wikipedia[Geometrical figures]
Plan: Search for more information about figures with four equal sides
#E2 = Wikipedia[Figures with four equal sides]
Plan: Find out the name of the figure with four equal sides, but no right angles
#E3 = LLM[What is the name of the figure with four equal sides, but no right angles? Given context: #E1 and #E2] | What name is given to a geometrical figure with four equal sides, but no right angles? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Rob Cavello.
#E1 = Wikipedia[Rob Cavello]
Plan: Find out which American rapper has worked with Rob Cavello.
#E2 = LLM[Which American rapper has worked with Rob Cavello? Given context: #E1] | Which American rapper from Ronkonkoma has worked with record producer Rob Cavello?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the 2009 Superbowl
#E1 = Wikipedia[2009 Superbowl]
Plan: Search for more information about the Pittsburgh Steelers
#E2 = Wikipedia[Pittsburgh Steelers]
Plan: Identify the losing team from the 2009 Superbowl
#E3 = LLM[Which team lost the 2009 Superbowl to the Pittsburgh Steelers? Given context: #E1 and #E2] | Which team lost the 2009 Superbowl to the Pittsburgh Steelers? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Rihanna's seventh studio album.
#E1 = Wikipedia[Rihanna's seventh studio album]
Plan: Find out the song that was the first time Rihanna and Guetta collaborated.
#E2 = LLM[What was the first time Rihanna and Guetta collaborated? Given context: #E1]
Plan: Search for more information about the song.
#E3 = Wikipedia[#E2] | "Phresh Out the Runway", is a song recorded by Barbadian recording artist Rihanna for her seventh studio album, and was the first time that Rihanna and Guetta had collaborated since which song by French DJ David Guetta from the reissue of his fourth studio album "One Love" (2009)?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the '.cat' top level domain
#E1 = Wikipedia['.cat' top level domain] | What does the internet top level domain '.cat' represent? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for books in the Bible
#E1 = Wikipedia[Books of the Bible]
Plan: Search for more information about the Four Horsemen of the Apocolypse
#E2 = Wikipedia[Four Horsemen of the Apocolypse]
Plan: Find out which book of the Bible describes the Four Horsemen of the Apocolypse and gives the 'number of the beast' as 666
#E3 = LLM[Which book of the Bible describes the Four Horsemen of the Apocolypse and gives the 'number of the beast' as 666? Given context: #E1 and #E2] | Which book of the Bible describes the 'Four Horsemen of the Apocolypse' and gives the 'number of the beast' as 666? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about fish
#E1 = Wikipedia[Fish]
Plan: Search for more information about fish that can hold objects
#E2 = Wikipedia[Fish that can hold objects] | Which is the only fish able to hold objects in its tail? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Borah Peak
#E1 = Wikipedia[Borah Peak]
Plan: Find out the name of the US state in which Borah Peak is located
#E2 = LLM[What is the name of the US state in which Borah Peak is located? Given context: #E1] | Borah Peak is the highest mountain in which US state? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Michael Snow.
#E1 = Wikipedia[Michael Snow]
Plan: Find out which country Michael Snow is from.
#E2 = LLM[Which country is Michael Snow from? Given context: #E1]
Plan: Search for more information about Wayne Wang.
#E3 = Wikipedia[Wayne Wang]
Plan: Find out which country Wayne Wang is from.
#E4 = LLM[Which country is Wayne Wang from? Given context: #E3] | Which artist is from Canada, Michael Snow or Wayne Wang?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the pineal gland
#E1 = Wikipedia[Pineal gland] | Where in the body is the pineal gland? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Alamut Castle
#E1 = Wikipedia[Alamut Castle]
Plan: Search for more information about the notorious group associated with Alamut Castle
#E2 = Wikipedia[notorious group associated with Alamut Castle] | Alamut Castle, now in modern-day Iran, is remembered today as the home of what notorious group between 1090 and 1256 AD? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Schweppes.
#E1 = Wikipedia[Schweppes]
Plan: Search for more information about Lilt.
#E2 = Wikipedia[Lilt]
Plan: Compare the countries in which Schweppes and Lilt are sold.
#E3 = LLM[Which beverage is sold in more countries, Schweppes or Lilt? Given context: #E1, #E2] | Which beverage is sold in more countries, Schweppes or Lilt?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Devil May Cry 3.
#E1 = Wikipedia[Devil May Cry 3]
Plan: Find out the year the game was released.
#E2 = LLM[In what year was Devil May Cry 3 released? Given context: #E1] | Devil May Cry 3: Dante's Awakening is a game that is a prequel to the game released in what year?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. |
Plan: Search for more information about jazz singing
#E1 = Wikipedia[jazz singing]
Plan: Identify the type of singing that is vocal improvisation with nonsense syllables or without words
#E2 = LLM[What type of singing is vocal improvisation with nonsense syllables or without words? Given context: #E1] | In jazz, what kind of singing is vocal improvisation with nonsense syllables or without words at all, whereby singers have the ability to create the equivalent of an instrumental solo using their voice? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Research about Dan Dare
#E1 = Wikipedia[Dan Dare]
Plan: Find out more information about Dan Dare's batman
#E2 = Wikipedia[Dan Dare's batman] | What is the name of Dan Dare’s Wigan-born batman? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the Greek alphabet
#E1 = Wikipedia[Greek alphabet]
Plan: Find out the next letter in the Greek alphabet
#E2 = LLM[Which is the next letter in the Greek alphabet? Given context: #E1] | Which is the next letter in the Greek alphabet – omicron, pi, rho, ______? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about The Leverhulme Trust.
#E1 = Wikipedia[The Leverhulme Trust]
Plan: Find out the name of the person who established The Leverhulme Trust.
#E2 = LLM[What is the name of the person who established The Leverhulme Trust? Given context: #E1]
Plan: Search for more information about the person.
#E3 = Wikipedia[#E2]
Plan: Find out the year in which the person was born.
#E4 = LLM[In what year was #E2 born? Given context: #E3] | In what year was the person born who's will established The Leverhulme Trust? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about The Lady Killer album.
#E1 = Wikipedia[The Lady Killer album]
Plan: Find out which song was parodied off of the album.
#E2 = LLM[What song did Whitney Avalon parody off of the 2010 The Lady Killer album? Given context: #E1] | What song did Whitney Avalon parody off of the 2010 "The Lady Killer" album?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the author.
#E1 = Wikipedia[Louisa May Alcott]
Plan: Find out what other novel the author is best known for.
#E2 = LLM[What other novel is the author Louisa May Alcott best known for? Given context: #E1] | A Long Fatal Love Chase was written by the author best known for what other novel?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the autobiography.
#E1 = Wikipedia["My Life as a 10-Year-Old Boy"]
Plan: Find out the name of the actress who wrote the autobiography.
#E2 = LLM[What is the name of the actress who wrote the autobiography "My Life as a 10-Year-Old Boy"? Given context: #E1] | Name the actress who wrote the autobiography "My Life as a 10-Year-Old Boy" who was also the voice of Bart Simpson and characters for the shows Todd Flanders and Database?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Claud A. Hatcher.
#E1 = Wikipedia[Claud A. Hatcher]
Plan: Find out the company that is engaged in the bottling and distribution in the Philippines of the cola-flavored soft drink developed by Claud A. Hatcher.
#E2 = LLM[Which company is engaged in the bottling and distribution in the Philippines of the cola-flavored soft drink developed by Claud A. Hatcher? Given context: #E1] | Which company is engaged in the bottling and distribution in the Philippines of the cola-flavored soft drink developed in 1905 by Claud A. Hatcher? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Patricia Kalember.
#E1 = Wikipedia[Patricia Kalember]
Plan: Find out what 1990 American psychological horror film directed by Adrian Lyne Patricia Kalember appeared in.
#E2 = LLM[What 1990 American psychological horror film directed by Adrian Lyne did Patricia Kalember appear in? Given context: #E1]
Plan: Find out who produced the film.
#E3 = LLM[Who produced the 1990 American psychological horror film directed by Adrian Lyne that Patricia Kalember appeared in? Given context: #E2] | The American actress "Patricia Kalember" appeared in this 1990 merican psychological horror film directed by Adrian Lyne, which was produced by whom? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Surrey Heath
#E1 = Wikipedia[Surrey Heath]
Plan: Find out the name of the MP for Surrey Heath
#E2 = LLM[Who is the MP for Surrey Heath? Given context: #E1]
Plan: Search for more information about the MP
#E3 = Wikipedia[name of the MP]
Plan: Find out the current role of the MP
#E4 = LLM[What is the current role of the MP? Given context: #E3] | Who is the MP for Surrey Heath and current Secretary of State for Education? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Noisia.
#E1 = Wikipedia[Noisia]
Plan: Find out the country of Noisia.
#E2 = LLM[What country is Noisia from? Given context: #E1] | I Am Legion is the self-titled collaborative album by "I Am Legion", the side project of Noisia, stylized as NOISIΛ ('VISION' rotated 180°), is an electronic music trio, of which country?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Danziger Echo.
#E1 = Wikipedia[Danziger Echo]
Plan: Find out the city state that Danziger Echo was published in.
#E2 = LLM[What is the name of the city state that Danziger Echo was published in? Given context: #E1]
Plan: Find out the years that the city state existed.
#E3 = LLM[Which years did the city state #E2 exist? Given context: #E1] | Danziger Echo was a newspaper that was published in the city state that existed between which years? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the Spanish conquistador
#E1 = Wikipedia[Hernan Cortes]
Plan: Search for more information about the conquest of Mexico
#E2 = Wikipedia[Conquest of Mexico] | What was the name of the Spanish conquistador who conquered Mexico in 1519? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the chemist.
#E1 = Wikipedia[chemist with highest Hirsch index rating in 2011]
Plan: Find out the name of the chemist's son.
#E2 = LLM[What is the name of the son of #E1? Given context: #E1] | What is the name of the son of the chemist who received the highest Hirsch index rating of all living chemists in 2011?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Kim Ji-won.
#E1 = Wikipedia[Kim Ji-won]
Plan: Find out which series Kim Ji-won starred in first.
#E2 = LLM[Which series did Kim Ji-won star in first, Gap-dong or The Heirs? Given context: #E1] | Which series did Kim Ji-won star in first, Gap-dong or The Heirs?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about counties in England
#E1 = Wikipedia[Counties in England]
Plan: Search for more information about bordering counties
#E2 = LLM[Which English county borders only one other county? Given context: #E1] | Which English county borders only one other county? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for information about the literary coterie
#E1 = Wikipedia[Blue Stockings Society]
Plan: Search for information about the members of the coterie
#E2 = Wikipedia[Mrs Vesey], Wikipedia[Mrs Ord], Wikipedia[Mrs Elizabeth Montagu], Wikipedia[Elizabeth Carter], Wikipedia[Mrs Hester Chapone], Wikipedia[Benjamin Stillingfleet] | What was the name of the literary coterie that met regularly around 1750 that included Mrs Vesey, Mrs Ord, Mrs Elizabeth Montagu, Elizabeth Carter, Mrs Hester Chapone and Benjamin Stillingfleet? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about General Motors X engine.
#E1 = Wikipedia[General Motors X engine]
Plan: Find out which engine shared bore diameter and some parts with General Motors X engine.
#E2 = LLM[Which engine shared bore diameter and some parts with General Motors X engine? Given context: #E1] | Designed by Chevrolet, this engine shared bore diameter and some parts with General Motors X engine.
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Malcolm X
#E1 = Wikipedia[Malcolm X]
Plan: Search for more information about his conversion to Islam
#E2 = Wikipedia[Malcolm X's conversion to Islam]
Plan: Find out the name Malcolm X had before converting to Islam.
#E3 = LLM[What was Malcolm X's surname before converting to Islam? Given context: #E1 and #E2] | What was Malcolm X's surname, before converting to Islam? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Glenn Miller
#E1 = Wikipedia[Glenn Miller]
Plan: Search for information about Glenn Miller's signature tune
#E2 = Wikipedia[Glenn Miller's signature tune] | What was Glenn Miller's signature tune? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Jay Silverheels
#E1 = Wikipedia[Jay Silverheels]
Plan: Search for more information about his most famous TV role
#E2 = LLM[What was the most famous TV role of actor Jay Silverheels? Given context: #E1] | What was the most famous TV role of actor Jay Silverheels |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Punic
#E1 = Wikipedia[Punic]
Plan: Identify which ancient city is associated with the Punic culture
#E2 = LLM[Which ancient city is associated with the Punic culture? Given context: #E1] | The adjective 'Punic' refers to which ancient city? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about The Dream of Gerontius
#E1 = Wikipedia[The Dream of Gerontius]
Plan: Search for more information about the oratorio
#E2 = Wikipedia[Gerontius oratorio]
Plan: Find out who set the poem to music in the oratorio
#E3 = LLM[Who set Cardinal Newman's poem The Dream of Gerontius to music in his 1900 oratorio? Given context: #E1 and #E2] | Who set Cardinal Newman's poem The Dream of Gerontius to music in his 1900 oratorio? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Kieron Moore.
#E1 = Wikipedia[Kieron Moore]
Plan: Find out which film adaptation of "Anna Karenina" Kieron Moore was in.
#E2 = LLM[Which film adaptation of "Anna Karenina" was Kieron Moore in? Given context: #E1]
Plan: Search for more information about the film adaptation.
#E3 = Wikipedia[#E2]
Plan: Find out which deceased English stage and film actress was opposite Kieron Moore in the film adaptation.
#E4 = LLM[Which deceased English stage and film actress was opposite Kieron Moore in the film adaptation #E2? Given context: #E3] | Kieron Moore, was an Irish film and television actor whose career was at its peak in the 1950s and 1960s, and may be best remembered for his role as Count Vronsky in the 1948 film adaptation of "Anna Karenina" opposite which deceased English stage and film actress?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Black Sabbath.
#E1 = Wikipedia[Black Sabbath]
Plan: Find out the album that featured the song "Changes".
#E2 = LLM[What album featured the song "Changes"? Given context: #E1]
Plan: Find out the band's manager when the album was released.
#E3 = LLM[Who was the band's manager when the album #E2 was released? Given context: #E1] | Who was the band's manager when Black Sabbath released the album that featured the song "Changes"?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Star Trek: The Experience.
#E1 = Wikipedia[Star Trek: The Experience]
Plan: Find out which Las Vegas location was scheduled to host the experience in 2009.
#E2 = LLM[Which Las Vegas location was scheduled to host Star Trek: The Experience in 2009? Given context: #E1] | What Las Vegas location, a $100 million entertainment complex containing 3 miles of neon lights, was scheduled to host Star Trek:The Experience in 2009? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Lost Songs from Lost Years.
#E1 = Wikipedia[Lost Songs from Lost Years]
Plan: Find out which state the indie rock band is from.
#E2 = LLM[Which state is the indie rock band from? Given context: #E1] | Lost Songs from Lost Years was released by an indie rock band from which state?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about cipher machines used during WWII
#E1 = Wikipedia[Cipher machines used during WWII]
Plan: Find out the name of the electro-mechanical cipher machine used by the Germans
#E2 = LLM[What is the name of the electro-mechanical cipher machine used by the Germans? Given context: #E1] | Consisting of rotors, a lampboard, a keyboard, and a plugboard, what is the name of the electro-mechanical cipher machine deployed by the Germans during WWII? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about J.P. Getty
#E1 = Wikipedia[J.P. Getty]
Plan: Ask LLM for the first names of J.P. Getty
#E2 = LLM[What were the first names of J.P. Getty? Given context: #E1] | What were the first names of the hugely wealthy J.P. Getty, 1932-2003? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the cochlea
#E1 = Wikipedia[Cochlea]
Plan: Search for more information about the stirrup
#E2 = Wikipedia[Stirrup]
Plan: Find out the location of the cochlea and the stirrup in the body
#E3 = LLM[Where in the body would you find the cochlea and the stirrup? Given context: #E1 and #E2] | Where in the body would you find the cochlea and the stirrup? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Mussolini
#E1 = Wikipedia[Mussolini]
Plan: Search for more information about Italy in October 1940
#E2 = Wikipedia[Italy in October 1940] | Ending humiliatingly for Italy, which country did Mussolini try to invade in October 1940? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Caipirinha
#E1 = Wikipedia[Caipirinha]
Plan: Search for more information about national cocktails
#E2 = Wikipedia[national cocktails]
Plan: Find out the South American country that has Caipirinha as its national cocktail.
#E3 = LLM[What is the South American country that has Caipirinha as its national cocktail? Given context: #E1 and #E2] | The drink Caipirinha is the national cocktail of which South American country? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Marlow
#E1 = Wikipedia[Marlow]
Plan: Find out the county in which Marlow is located
#E2 = LLM[What county is Marlow located in? Given context: #E1] | In which county is the Thames-side town of Marlow? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Research the 1961 film ‘West Side Story’
#E1 = Wikipedia[West Side Story (1961 film)]
Plan: Identify the rival gangs in the film
#E2 = LLM[Who are the rival gangs in the 1961 film 'West Side Story'? Given context: #E1] | Who are the rival gangs in the 1961 film ‘West Side Story’? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about bull-leaping
#E1 = Wikipedia[Bull-leaping]
Plan: Search for more information about Mediterranean islands
#E2 = Wikipedia[Mediterranean islands]
Plan: Find out which Mediterranean island the ancient ritual of bull-leaping took place on.
#E3 = LLM[Which Mediterranean island did the ancient ritual of bull-leaping take place on? Given context: #E1 and #E2] | The ancient ritual of bull-leaping took place on which Mediterranean island? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about J.L. Austin.
#E1 = Wikipedia[J.L. Austin]
Plan: Find out about the speech acts described by J.L. Austin.
#E2 = LLM[What are some classes of speech acts described by J.L. Austin? Given context: #E1] | What is one class of speech acts described by J.L. Austin?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the British Commonwealth
#E1 = Wikipedia[British Commonwealth]
Plan: Search for more information about countries that left the British Commonwealth
#E2 = Wikipedia[countries that left the British Commonwealth] | Which country left the British Commonwealth in 1949? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Feroze Gandhi.
#E1 = Wikipedia[Feroze Gandhi]
Plan: Find out the name of the wife of Feroze Gandhi.
#E2 = LLM[What is the name of the wife of Feroze Gandhi? Given context: #E1]
Plan: Search for more information about the wife of Feroze Gandhi.
#E3 = Wikipedia[#E2]
Plan: Find out to which political party the wife of Feroze Gandhi belonged.
#E4 = LLM[To which political party did the wife of Feroze Gandhi belong? Given context: #E3] | To what political party did the wife of Feroze Gandhi belong ? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. |
Plan: Search for more information about the Seattle Seahawks
#E1 = Wikipedia[Seattle Seahawks]
Plan: Search for more information about the 1976 NFL season
#E2 = Wikipedia[1976 NFL season]
Plan: Find out which expansion team the Seattle Seahawks beat on Oct 17, 1976
#E3 = LLM[Which expansion team did the Seattle Seahawks beat on Oct 17, 1976? Given context: #E1 and #E2] | On Oct 17, 1976, the Seattle Seahawks won their first ever game by beating which other expansion team? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Ford car models
#E1 = Wikipedia[Ford car models]
Plan: Search for more information about Italian alpine resorts
#E2 = Wikipedia[Italian alpine resorts]
Plan: Find out the name of the Ford car model that takes its name from an Italian alpine resort.
#E3 = LLM[What is the name of the Ford car model that takes its name from an Italian alpine resort? Given context: #E1 and #E2] | Which model of Ford car took its name from an Italian alpine resort? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the single.
#E1 = Wikipedia[Love Drunk]
Plan: Find out which American singer features the single.
#E2 = LLM[Which American singer features the single Love Drunk? Given context: #E1] | Love Drunk included the single that features which American singer?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about King of the Grizzlies.
#E1 = Wikipedia[King of the Grizzlies]
Plan: Search for more information about D2: The Mighty Ducks.
#E2 = Wikipedia[D2: The Mighty Ducks]
Plan: Compare the two and determine which organization released them.
#E3 = LLM[Which organization released King of the Grizzlies and D2: The Mighty Ducks? Given context: #E1, #E2] | King of the Grizzlies and D2: The Mighty Ducks, were released by the same organization? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the philosopher.
#E1 = Wikipedia[German-speaking philosopher]
Plan: Find out the birthplace of the philosopher.
#E2 = LLM[What is the birthplace of the German-speaking philosopher? Given context: #E1]
Plan: Search for more information about the birthplace.
#E3 = Wikipedia[#E2]
Plan: Find out if the philosopher was born of Jewish parentage.
#E4 = LLM[Was the German-speaking philosopher born of Jewish parentage? Given context: #E3]
Plan: Find out if the philosopher praised a Polish rabbi for his benevolence.
#E5 = LLM[Did the German-speaking philosopher praise a Polish rabbi for his benevolence? Given context: #E4] | What German-speaking philosopher, born of Jewish parentage in present-day ,praises a Polish rabbi for his benevolence?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about The Undertones
#E1 = Wikipedia[The Undertones]
Plan: Search for more information about the lead singer
#E2 = Wikipedia[The Undertones lead singer] | The lead singer of The Undertones later achieved success as a solo act, what was his name? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Arkham Asylum
#E1 = Wikipedia[Arkham Asylum]
Plan: Search for more information about famous residents
#E2 = Wikipedia[Notable inmates of Arkham Asylum] | Who is Arkham Asylum's most famous resident? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Research more information about Lorenzo the Magnificent
#E1 = Wikipedia[Lorenzo the Magnificent]
Plan: Find out what city-state Lorenzo the Magnificent ruled
#E2 = LLM[What city-state did Lorenzo the Magnificent rule? Given context: #E1] | Lorenzo the Magnificent was the ruler of what city-state? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the composer
#E1 = Wikipedia["A Child of our Time"]
Plan: Find out the name of the composer of the oratorio
#E2 = LLM[Who composed the oratorio "A Child of our Time"? Given context: #E1] | "Which composer wrote the 1944 oratorio ""A Child of our Time""?" |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the film.
#E1 = Wikipedia[Joe Wright film]
Plan: Find out what novel the film is based on.
#E2 = LLM[What is the name of the novel that Joe Wright's 2012 film is based on? Given context: #E1]
Plan: Search for more information about the novel.
#E3 = Wikipedia[#E2] | Who wrote the 1877 novel which Joe Wright adopted into a 2012 film?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Elgin in Moray
#E1 = Wikipedia[Elgin in Moray]
Plan: Find out the name of the school in Elgin, Moray
#E2 = LLM[What is the name of the school in Elgin, Moray? Given context: #E1] | Which renowned independent school is situated in Elgin in Moray ? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Henry 'Hotspur' Percy
#E1 = Wikipedia[Henry 'Hotspur' Percy]
Plan: Search for more information about the battle of 1403
#E2 = Wikipedia[Battle of Shrewsbury] | At which battle of 1403 did Henry 'Hotspur' Percy die? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Bob Sawyer
#E1 = Wikipedia[Bob Sawyer]
Plan: Search for more information about Charles Dickens novels
#E2 = Wikipedia[Charles Dickens Novels]
Plan: Find out which Dickens novel Bob Sawyer appears in
#E3 = LLM[Which Dickens novel does Bob Sawyer appear in? Given context: #E1 and #E2] | In which Dickens novel does Bob Sawyer appear? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the ITV series Ivanhoe
#E1 = Wikipedia[Ivanhoe (TV series)]
Plan: Search for more information about the actor who played the title role
#E2 = Wikipedia[List of Ivanhoe (TV series) cast members] | Who played the title role the ITV series Ivanhoe, first broadcast in 1958? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Baar railway station.
#E1 = Wikipedia[Baar railway station]
Plan: Find out the municipality and canton of the Baar railway station.
#E2 = LLM[What is the municipality and canton of Baar railway station? Given context: #E1] | Baar railway station is situated in this municipality in the canton of what
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Shinedown.
#E1 = Wikipedia[Shinedown]
Plan: Find out when Shinedown was formed.
#E2 = LLM[When was Shinedown formed? Given context: #E1]
Plan: Search for more information about The dB's.
#E3 = Wikipedia[The dB's]
Plan: Find out when The dB's was formed.
#E4 = LLM[When was The dB's formed? Given context: #E3] | What band formed first, Shinedown or The dB's?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Arthur's Seat
#E1 = Wikipedia[Arthur's Seat]
Plan: Search for more information about cities in Scotland
#E2 = Wikipedia[Cities in Scotland]
Plan: Find out which city Arthur's Seat is located in.
#E3 = LLM[Which city is Arthur's Seat located in? Given context: #E1 and #E2] | In which Scottish city is Arthur's Seat? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about A Thousand Winding Roads.
#E1 = Wikipedia[A Thousand Winding Roads]
Plan: Find out which 1993 debut album the song was recorded for.
#E2 = LLM[Which 1993 debut album did A Thousand Winding Roads include a song that was later recorded for? Given context: #E1] | A Thousand Winding Roads included a song that would later be recorded for which 1993 debut album?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Nick Grinde.
#E1 = Wikipedia[Nick Grinde]
Plan: Find out Nick Grinde's nationality.
#E2 = LLM[What is Nick Grinde's nationality? Given context: #E1]
Plan: Search for more information about Andy Tennant.
#E3 = Wikipedia[Andy Tennant]
Plan: Find out Andy Tennant's nationality.
#E4 = LLM[What is Andy Tennant's nationality? Given context: #E3]
Plan: Compare Nick Grinde's nationality and Andy Tennant's nationality.
#E5 = LLM[Do Nick Grinde and Andy Tennant share the same nationality? Given context: #E2, #E4] | Do Nick Grinde and Andy Tennant share the same nationality?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about weather forecasting
#E1 = Wikipedia[Weather forecasting]
Plan: Find out the term applied to the process of gathering together weather forecasts from various recording stations
#E2 = LLM[What is the term applied to the process of gathering together weather forecasts from various recording stations? Given context: #E1] | What is the term applied to the process of gathering together weather forecasts from various recording stations? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Mike Muir.
#E1 = Wikipedia[Mike Muir]
Plan: Search for more information about Leo.
#E2 = Wikipedia[Leo]
Plan: Compare the two and determine which singer is South Korean.
#E3 = LLM[ | Which singer is South Korean, Mike Muir or Leo? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Melbourne Museum
#E1 = Wikipedia[Melbourne Museum]
Plan: Search for more information about race horses
#E2 = Wikipedia[race horses]
Plan: Find out the name of the race horse whose mounted hide is displayed at the Melbourne Museum.
#E3 = LLM[What is the name of the race horse whose mounted hide is displayed at the Melbourne Museum? Given context: #E1 and #E2] | The mounted hide of which race horse is displayed at the Melbourne Museum in Australia? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about hyperborean
#E1 = Wikipedia[hyperborean]
Plan: Search for more information about the destination
#E2 = LLM[Where do I end up if I travel in a hyperborean direction as far as I can go? Given context: #E1] | Travel in a hyperborean direction as far as you can go. Where do you end up? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about ornamental designs
#E1 = Wikipedia[ornamental design]
Plan: Search for more information about the material used for ornamental designs
#E2 = Wikipedia[hard material used for ornamental designs]
Plan: Find out the name for an ornamental design made of small pieces of hard material
#E3 = LLM[What is the name for an ornamental design made of small pieces of hard material? Given context: #E1 and #E2] | What name is given to an ornamental, usually coloured, design made of small pieces of hard material such as stone, tile or glass? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the battle
#E1 = Wikipedia[Battle of Marston Moor]
Plan: Search for the location of the battle
#E2 = Wikipedia[Marston Moor]
Plan: Find out the name of the place between York and Knaresborough in the UK where the battle of Marston Moor took place.
#E3 = LLM[What is the name of the place between York and Knaresborough in the UK where the battle of Marston Moor took place? Given context: #E1 and #E2] | Which place between York and Knaresborough in the UK was the site of a decisive battle in 1642 when Cromwell defeated Royalist forces? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the Temple of Confucius
#E1 = Wikipedia[Temple of Confucius]
Plan: Find out the location of the Temple of Confucius
#E2 = LLM[What is the location of the Temple of Confucius? Given context: #E1] | Where is the Temple of Confucius? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Sempervivum.
#E1 = Wikipedia[Sempervivum]
Plan: Find out how many species are in the genus Sempervivum.
#E2 = LLM[How many species are in the genus Sempervivum? Given context: #E1]
Plan: Search for more information about Malope.
#E3 = Wikipedia[Malope]
Plan: Find out how many species are in the genus Malope.
#E4 = LLM[How many species are in the genus Malope? Given context: #E3]
Plan: Compare the number of species in both genera.
#E5 = LLM[Which genus has more species between Sempervivum and Malope? Given context: #E2, #E4] | Between Sempervivum and Malope, which genus has more species?
|
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about photographers who take pictures of crowds of naked people
#E1 = Wikipedia[Photographer who take pictures of crowds of naked people]
Plan: Identify the photographer from the search results
#E2 = LLM[Who is the photographer famous for taking pictures of crowds of naked people? Given context: #E1] | What is the name of the photographer famous for taking pictures of crowds of naked people? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the actor who starred in the adventure drama "Extreme".
#E1 = Wikipedia["Extreme" (film)]
Plan: Find out the name of the actor who starred in the film.
#E2 = LLM[What is the name of the actor who starred in the film "Extreme"? Given context: #E1]
Plan: Search for more information about the actor.
#E3 = Wikipedia[#E2]
Plan: Find out the year the actor was born.
#E4 = LLM[In what year was #E2 born? Given context: #E3] | In what year was the actor who starred in the adventure drama "Extreme" born? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about Pawe Ciompa
#E1 = Wikipedia[Pawe Ciompa]
Plan: Search for more information about Ragnar Frisch
#E2 = Wikipedia[Ragnar Frisch]
Plan: Find out the word invented by Pawe Ciompa and first used by Ragnar Frisch in the sense that it is used today.
#E3 = LLM[What word, invented by Pawe Ciompa and first used by Ragnar Frisch in the sense that it is used today, is concerned with developing and applying quantitative or statistical methods to the study and elucidation of economic principles, combining economic theory with statistics to analyse and test economic relationships? Given context: #E1 and #E2] | What word, invented by Pawe Ciompa in 1910 and first used by Ragnar Frisch in the sense that it is used today, is concerned with developing and applying quantitative or statistical methods to the study and elucidation of economic principles, combining economic theory with statistics to analyse and test economic relationships? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the song
#E1 = Wikipedia["Fanfare for the Common Man" (song)]
Plan: Find out which group had a 1977 No 2 UK hit with the song
#E2 = LLM[Which group had a 1977 No 2 UK hit with the song "Fanfare for the Common Man"? Given context: #E1] | Which group had a 1977 No 2 UK hit with a rock instrumental 'Fanfare for the Common Man'? |
For the following tasks, make plans that can solve the problem step-by-step. For each plan, indicate which external tool together with tool input to retrieve evidence. You can store the evidence into a variable #E that can be called by later tools. (Plan, #E1, Plan, #E2, Plan, ...)
Tools can be one of the following:
Wikipedia[input]: Worker that search for similar page contents from Wikipedia. Useful when you need to get holistic knowledge about people, places, companies, historical events, or other subjects. The response are long and might contain some irrelevant information. Input should be a search query.
LLM[input]: A pretrained LLM like yourself. Useful when you need to act with general world knowledge and common sense. Prioritize it when you are confident in solving the problem yourself. Input can be any instruction. | Plan: Search for more information about the Indian Tamil language film, VIP.
#E1 = Wikipedia[VIP (2017 film)]
Plan: Find out who managed cinematography for the film.
#E2 = LLM[Who managed cinematography for the Indian Tamil language film VIP? Given context: #E1] | Who managed cinematography for the Indian Tamil language film, VIP, whose music was composed by longtime associate of A R Rahman, Ranjit Barot? |
End of preview. Expand
in Dataset Viewer.
README.md exists but content is empty.
Use the Edit dataset card button to edit it.
- Downloads last month
- 82