Dataset Viewer
Auto-converted to Parquet Duplicate
context
listlengths
1
1
gen_kwargs
dict
repeat_idx
int64
0
4
request_idx
int64
0
499
task_name
stringclasses
3 values
metadata
dict
model_outputs
stringlengths
745
186k
[ { "content": "You are a math expert. You will be given a math problem. Please solve it step by step and provide your final answer as an integer. Your response MUST contain the final answer inside '\\boxed{}', like '\\boxed{82}', otherwise it will be considered incorrect automatically. \n\n Question: Among the 9...
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
0
AIME24
{ "expected_answer": "73", "problem_id": "aime24-0", "reference_solution": "Let $w,x,y,z$ denote the number of residents who own 1,2,3 and 4 of these items, respectively. We know $w+x+y+z=900$ , since there are 900 residents in total. This simplifies to $w+z=229$ , since we know $x=437$ and $y=234$ . Now, w...
Alright, let's tackle this problem step by step. We have 900 residents in Aimeville, and each of them owns a bag of candy hearts. Additionally, they might own a diamond ring, a set of golf clubs, or a garden spade. We need to find out how many residents own all four of these things. However, since they already own a ba...
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
1
AIME24
{"expected_answer":"236","problem_id":"aime24-1","reference_solution":"The third condition implies t(...TRUNCATED)
"Alright, let's tackle this problem step by step. So, we have a list of positive integers with three(...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
2
AIME24
{"expected_answer":"45","problem_id":"aime24-2","reference_solution":"Consider this table: $\\begin{(...TRUNCATED)
"Alright, let's tackle this problem step by step. We have a 2x3 grid where each cell can hold a digi(...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
3
AIME24
{"expected_answer":"33","problem_id":"aime24-3","reference_solution":"Denote $\\log_2(x) = a$ , $\(...TRUNCATED)
"Alright, let's tackle this problem step by step. We have three equations involving logarithms, and (...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
4
AIME24
{"expected_answer":"80","problem_id":"aime24-4","reference_solution":"Draw a good diagram!\nLet $AF(...TRUNCATED)
"Alright, let's tackle this problem step by step. So, we have a convex equilateral hexagon ABCDEF wh(...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
5
AIME24
{"expected_answer":"55","problem_id":"aime24-5","reference_solution":"Let $k$ be one of the elemen(...TRUNCATED)
"Alright, let's tackle this problem step by step. So Alice chooses a set A of positive integers, and(...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
6
AIME24
{"expected_answer":"699","problem_id":"aime24-6","reference_solution":"We note that by changing a di(...TRUNCATED)
"Alright, let's tackle this problem step by step. So, we need to find the greatest four-digit positi(...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
7
AIME24
{"expected_answer":"127","problem_id":"aime24-7","reference_solution":"First, let's consider a secti(...TRUNCATED)
"Alright, let's tackle this problem step by step. So, we have a torus (a donut-shaped surface) and a(...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
8
AIME24
{"expected_answer":"902","problem_id":"aime24-8","reference_solution":"The problem says \"some\", so(...TRUNCATED)
"Alright, let's tackle this problem step by step. So, we have a 5x5 grid with 25 unit cells, and we (...TRUNCATED)
[{"content":"You are a math expert. You will be given a math problem. Please solve it step by step a(...TRUNCATED)
{ "do_sample": false, "max_new_tokens": 32768, "seed": 0, "temperature": 0.7 }
0
9
AIME24
{"expected_answer":"468","problem_id":"aime24-9","reference_solution":"By Euler's formula $OI^{2}=R(...TRUNCATED)
"Alright, let's tackle this problem step by step. We have a triangle ABC with circumcenter O and inc(...TRUNCATED)
End of preview. Expand in Data Studio

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

mlfoundations-dev/s1K-sharegpt_1743202401_eval_0771

Precomputed model outputs for evaluation.

Evaluation Results

Summary

Metric AIME24 AIME25 MATH500
Accuracy 14.0 8.7 54.4

AIME24

  • Average Accuracy: 14.00% ± 1.12%
  • Number of Runs: 5
Run Accuracy Questions Solved Total Questions
1 13.33% 4 30
2 16.67% 5 30
3 13.33% 4 30
4 16.67% 5 30
5 10.00% 3 30

AIME25

  • Average Accuracy: 8.67% ± 1.52%
  • Number of Runs: 5
Run Accuracy Questions Solved Total Questions
1 3.33% 1 30
2 10.00% 3 30
3 6.67% 2 30
4 13.33% 4 30
5 10.00% 3 30

MATH500

  • Accuracy: 54.40%
    Accuracy Questions Solved Total Questions
    54.40% 272 500
Downloads last month
4