prompt stringlengths 2.51k 289k | response stringclasses 88
values | candidate_agents listlengths 1 11 | data_source stringclasses 1
value | ability stringclasses 1
value | label stringclasses 3
values | agent_names listlengths 0 3 | agent_name stringclasses 37
values | num_agents int64 0 3 | target_error_type null | trajectory_id null | id stringclasses 12
values | oversampled bool 1
class | oversample_repeat int64 1 1 | oversample_min_agent_freq null | oversample_row_id null |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"A","agents":["RoleAssigner"]} | [
"RoleAssigner",
"Solver",
"Evaluator"
] | maserror_nli_verifier | nli_verification | A | [
"RoleAssigner"
] | RoleAssigner | 1 | null | null | maserror/agent_detection_nli_1 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"C","agents":[]} | [
"RoleAssigner",
"Solver",
"Evaluator"
] | maserror_nli_verifier | nli_verification | C | [] | null | 0 | null | null | maserror/agent_detection_nli_2 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"B","agents":[]} | [
"RoleAssigner",
"Solver",
"Evaluator"
] | maserror_nli_verifier | nli_verification | B | [] | null | 0 | null | null | maserror/agent_detection_nli_3 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"A","agents":["Solver"]} | [
"RoleAssigner",
"Solver",
"Evaluator"
] | maserror_nli_verifier | nli_verification | A | [
"Solver"
] | Solver | 1 | null | null | maserror/agent_detection_nli_4 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"A","agents":["Evaluator"]} | [
"RoleAssigner",
"Solver",
"Evaluator"
] | maserror_nli_verifier | nli_verification | A | [
"Evaluator"
] | Evaluator | 1 | null | null | maserror/agent_detection_nli_7 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"C","agents":[]} | [
"RoleAssigner",
"Solver",
"Evaluator"
] | maserror_nli_verifier | nli_verification | C | [] | null | 0 | null | null | maserror/agent_detection_nli_8 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"A","agents":["assistant2"]} | [
"Assistant1",
"Assistant2",
"Assistant3",
"assistant2"
] | maserror_nli_verifier | nli_verification | A | [
"assistant2"
] | assistant2 | 1 | null | null | maserror/agent_detection_nli_1 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"B","agents":[]} | [
"Assistant1",
"Assistant2",
"Assistant3",
"assistant2"
] | maserror_nli_verifier | nli_verification | B | [] | null | 0 | null | null | maserror/agent_detection_nli_2 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"A","agents":["assistant4"]}
{"label":"A","agents":["assistant3"]} | [
"Assistant1",
"Assistant2",
"Assistant3",
"assistant4",
"assistant3"
] | maserror_nli_verifier | nli_verification | A | [
"assistant4",
"assistant3"
] | assistant4 | 2 | null | null | maserror/agent_detection_nli_4 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"B","agents":[]} | [
"Assistant1",
"Assistant2",
"Assistant3",
"assistant4",
"assistant3"
] | maserror_nli_verifier | nli_verification | B | [] | null | 0 | null | null | maserror/agent_detection_nli_5 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"A","agents":["ranker5"]} | [
"Assistant1",
"Assistant2",
"Assistant3",
"ranker5"
] | maserror_nli_verifier | nli_verification | A | [
"ranker5"
] | ranker5 | 1 | null | null | maserror/agent_detection_nli_7 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"B","agents":[]} | [
"Assistant1",
"Assistant2",
"Assistant3",
"ranker5"
] | maserror_nli_verifier | nli_verification | B | [] | null | 0 | null | null | maserror/agent_detection_nli_8 | false | 1 | null | null |
You are a careful verifier for multi-agent trajectory failure attribution.
You will be given:
1. A trajectory
2. A failure hypothesis
3. The list of candidate agents appearing in the trajectory
Your task is to jointly predict:
- label: one of A, B, C
- agents: the responsible agent list for this JSON object
Label me... | {"label":"A","agents":["RoleAssigner"]}
{"label":"A","agents":["Evaluator"]} | [
"RoleAssigner",
"Solver",
"Critic",
"Evaluator"
] | maserror_nli_verifier | nli_verification | A | [
"RoleAssigner",
"Evaluator"
] | RoleAssigner | 2 | null | null | maserror/agent_detection_nli_1 | false | 1 | null | null |
End of preview. Expand in Data Studio
No dataset card yet
- Downloads last month
- 33