Hugging Face's logo

Datasets:
Yifan99
/
BenchFill

Tasks:
Question Answering
Modalities:
Text
Formats:
json
Languages:
English
Size:
< 1K
ArXiv:
Tags:
mathematics
benchmark
evaluation
benchfill
Libraries:
Datasets
pandas
Polars
License:
Dataset card Data Studio Files
xet
 Community
1
BenchFill / README.md
Yifan99's picture
Yifan99
Update citation to the published arXiv record
70cce22 verified
preview code
|
Raw History Blame Contribute Delete
3.96 kB
metadata 
license: apache-2.0
language:
  - en
task_categories:
  - question-answering
tags:
  - mathematics
  - benchmark
  - evaluation
  - benchfill
size_categories:
  - n<1K
configs:
  - config_name: AIME24
    data_files:
      - split: test
        path: AIME24.jsonl
  - config_name: AIME25
    data_files:
      - split: test
        path: AIME25.jsonl
  - config_name: AIME26
    data_files:
      - split: test
        path: AIME26.jsonl
  - config_name: AMC23
    data_files:
      - split: test
        path: AMC23.jsonl

BenchFill

Four augmented mathematics evaluation benchmarks from Your Benchmark Is an Empirical Measure Over Difficulty (NeurIPS 2026), by Yifan Sun, Naicheng Yu, Jiawen Gong, Jingyan Shen, and Huan Zhang.

BenchFill generates new items to fill gaps in a benchmark's difficulty distribution. This release contains the original benchmark items followed by the curated generated items.

Why BenchFill?

BenchFill helps detect training gains that existing benchmarks miss. Filling gaps in benchmark difficulty makes localized improvements easier to observe. In the paper's SFT experiment, adding just one BenchFill question to each of three AIME benchmarks made training gains statistically detectable in all three cases; the original benchmarks and baseline augmentations detected none (Table 12).

How BenchFill works: training gains in a difficulty gap may be missed by the original benchmark; adding questions in that gap makes the gains easier to detect.

Subsets

All four configurations use the test split.

Configuration Original items Generated items Total Original JSON Hugging Face JSONL
AIME24 30 20 50 AIME24.json AIME24.jsonl
AIME25 30 22 52 AIME25.json AIME25.jsonl
AIME26 30 23 53 AIME26.json AIME26.jsonl
AMC23 40 40 80 AMC23.json AMC23.jsonl
Total 130 105 235

Usage 

from datasets import load_dataset

dataset = load_dataset("Yifan99/BenchFill", "AIME24", split="test")
# Other configurations: "AIME25", "AIME26", "AMC23"

original = dataset.filter(lambda row: "-origin-" in row["id"])
generated = dataset.filter(lambda row: "-fill-" in row["id"])

Data format

Each record has three fields:

  • id: question ID.
  • question: mathematics problem text.
  • answer: reference answer.

IDs use -origin- for original questions and -fill- for generated questions. Both JSON and JSONL files are provided; load_dataset uses JSONL.

Source attribution and terms

The repository's existing Apache-2.0 metadata is retained. Original AIME and AMC competition materials are from the Mathematical Association of America (MAA) and retain their original source terms; the repository metadata does not relicense those materials.

See the upstream THIRD_PARTY_NOTICES.md, including its AMC23 and AIME24/25/26 attribution and MAA educational-sharing notice. Source notices and permissions apply to the corresponding source content.

Citation 

@misc{sun2026benchmarkempiricalmeasuredifficulty,
  title = {Your Benchmark Is an Empirical Measure Over Difficulty},
  author = {Yifan Sun and Naicheng Yu and Jiawen Gong and Jingyan Shen and Huan Zhang},
  year = {2026},
  eprint = {2610.29927},
  archivePrefix = {arXiv},
  primaryClass = {cs.LG},
  url = {https://arxiv.org/abs/2610.29927}
}