datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
math_difficulty_datadapo14k_difficultyRL-seed-Decensor-Difficultymath_stratos_scale_judged_and_annotated_with_difficultyAM-Math-Difficulty-RLFor more open-source datasets, models, and methodologies, please visit our GitHub repository.
We believe that the selection of training data for reinforcement learning is crucial.
To validate this, we conducted several experiments exploring how data difficulty influences training performance.
Our data sources originate from numerous excellent open-source projects, and we sincerely appreciate their contributions, without which our current achievements would not have been possible.… See the full description on the dataset page: https://huggingface.co/datasets/a-m-team/AM-Math-Difficulty-RL.prompt-difficulty
Prompt Difficulty Assessment
Prompt difficulty plays a critical role in the performance of large language models (LLMs).
Assessing this difficulty is essential for selecting training examples, evaluating model capabilities, and optimizing routing and reasoning strategies.
Yet, no standardized framework exists for comparing prompt difficulty across domains.
This report proposes a method to quantify prompt difficulty using multiple LLMs and introduces a composite difficulty score for… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/prompt-difficulty.reeval-difficulty-for-helmMMMU_with_difficulty_level
MMMU with difficulty level tags
This dataset extends the 🤗 MMMU val benchmark by introducing two additional tags: passrate_for_qwen2.5_vl_7b and difficulty_level_for_qwen2.5_vl_7b. Further details are available in our paper The Synergy Dilemma of Long-CoT SFT and RL: Investigating Post-Training Techniques for Reasoning VLMs.
🚀 Data Usage
from datasets import load_dataset
dataset = load_dataset("JierunChen/MMMU_with_difficulty_level")
print(dataset)
📑… See the full description on the dataset page: https://huggingface.co/datasets/JierunChen/MMMU_with_difficulty_level.DeepScaleR_Difficulty
Difficulty Estimation on DeepScaleR
We annotate the entire DeepScaleR dataset with a difficulty score based on the performance of the Qwen 2.5-MATH-7B model. This provides an adaptive signal for curriculum construction and model evaluation.
DeepScaleR is a curated dataset of 40,000 reasoning-intensive problems used to train and evaluate reinforcement learning-based methods for large language models.
Difficulty Scoring Method
Difficulty scores are estimated using the… See the full description on the dataset page: https://huggingface.co/datasets/lime-nlp/DeepScaleR_Difficulty.Exercise-Synthetic-split-ncert-chapter-mapped_filtered_difficulty_scoredGSM8K_Difficulty
Difficulty Estimation on DeepScaleR
We annotate the entire GSM8K dataset with a difficulty score based on the performance of the Qwen 2.5-MATH-7B model. This provides an adaptive signal for curriculum construction and model evaluation.
GSM8K (Grade School Math 8K) is a dataset of 8.5K high quality linguistically diverse grade school math word problems. The dataset was created to support the task of question answering on basic mathematical problems that require multi-step reasoning.… See the full description on the dataset page: https://huggingface.co/datasets/lime-nlp/GSM8K_Difficulty.aozora-text-difficulty
Aozora Text Difficulty Dataset
This dataset contains Japanese literary texts from the Aozora Bunko digital library, enhanced with jReadability-based difficulty analysis for Japanese language learning and curriculum development.
Dataset Overview
Source: Aozora Bunko (青空文庫) - Japan's premier digital library of public domain literature
Enhancement: jReadability-based difficulty scoring using research-backed Japanese readability models
Primary Methodology: jReadability - A… See the full description on the dataset page: https://huggingface.co/datasets/ronantakizawa/aozora-text-difficulty.BlindLoop-Difficulty-Feedback
BlindLoop Difficulty Feedback
This is the public, hash-bound release of BlindLoop Section 3. Coding agents
generated executable visual-question tasks; each task's inverse program checked
the answer from rendered pixels. For complete feedback transactions, the exact
same five images were evaluated by three frontier VLMs and the resulting
difficulty signal was returned to the next generation episode.
Contents
Config
Unit
Rows
tasks
generated task
266… See the full description on the dataset page: https://huggingface.co/datasets/taesiri/BlindLoop-Difficulty-Feedback.seed_math_exploit_difficulty_annotationdifficulty-E2H-AMC-generations
Generations Dataset: E2H-AMC
LLM-generated solutions across train/validation/test splits for multiple models.
Columns
Column
Type
Description
problem
str
Problem statement
generated_solutions
list
Generated solutions with scores
success_rate
float
Fraction of correct generations
majority_vote_is_correct
int (0/1)
Whether majority vote is correct
k
int
Number of samples generated
temperature
float
Sampling temperature
max_len
int
Maximum… See the full description on the dataset page: https://huggingface.co/datasets/CoffeeGitta/difficulty-E2H-AMC-generations.Skywork-DAPOCL180-difficultyMathVision_with_difficulty_level
MathVision with difficulty level tags
This dataset extends the 🤗 MathVision benchmark by introducing two additional tags: passrate_for_qwen2.5_vl_7b and difficulty_level_for_qwen2.5_vl_7b. Further details are available in our paper The Synergy Dilemma of Long-CoT SFT and RL: Investigating Post-Training Techniques for Reasoning VLMs.
🚀 Data Usage
from datasets import load_dataset
dataset = load_dataset("JierunChen/MathVision_with_difficulty_level")
print(dataset)… See the full description on the dataset page: https://huggingface.co/datasets/JierunChen/MathVision_with_difficulty_level.difficulty-gsm8k-generations
Generations Dataset: gsm8k
LLM-generated solutions across train/validation/test splits for multiple models.
Columns
Column
Type
Description
problem
str
Problem statement
generated_solutions
list
Generated solutions with scores
success_rate
float
Fraction of correct generations
majority_vote_is_correct
int (0/1)
Whether majority vote is correct
k
int
Number of samples generated
temperature
float
Sampling temperature
max_len
int
Maximum… See the full description on the dataset page: https://huggingface.co/datasets/CoffeeGitta/difficulty-gsm8k-generations.Qwen2.5-7B-Instruct-Turbo_labeled_numina_difficulty_162K_10difficulty-eval-64difficulty-aime_2025-generations
Generations Dataset: aime_2025
Paper: LLMs Encode Their Failures: Predicting Success from Pre-Generation ActivationsCode: GitHub
LLM-generated solutions across train/validation/test splits for multiple models.
Columns
Column
Type
Description
problem
str
Problem statement
generated_solutions
list
Generated solutions with scores
success_rate
float
Fraction of correct generations
majority_vote_is_correct
int (0/1)
Whether majority vote is correct
k… See the full description on the dataset page: https://huggingface.co/datasets/CoffeeGitta/difficulty-aime_2025-generations.bible-vocabulary-difficulty
Bible vocabulary-difficulty metrics, 12 translations
Per-verse reading-difficulty metrics plus a cross-language book-name table
keyed on USFM codes. Produced by bible-reader — see
scripts/export_dataset.py.
No verse text
This dataset contains references and derived metrics only, never verse
text. That is deliberate: it keeps translations under copyright (NASB)
publishable as derived data, and it keeps the download small. Fetch the texts
themselves from their own… See the full description on the dataset page: https://huggingface.co/datasets/lego573402/bible-vocabulary-difficulty.dapo-math-17k-difficulty-qwen3-1.7b-base-k16
DAPO-Math-17k difficulty under Qwen3-1.7B-Base (K=16)
For each of the 17,398 problems in the DAPO-Math-17k train set, how many of
K=16 samples from the untrained base model are correct.
The headline: 57.27% of problems are solved 0 out of 16 times, and not one
problem is solved 16 out of 16. Difficulty here is entirely one-sided.
Why count per problem instead of reporting mean accuracy
In group-relative RL (GRPO and its relatives), a prompt group whose K responses… See the full description on the dataset page: https://huggingface.co/datasets/RyanYr/dapo-math-17k-difficulty-qwen3-1.7b-base-k16.MathVista_with_difficulty_level
MathVista with difficulty level tags
This dataset extends the 🤗 MathVista testmini benchmark by introducing two additional tags: passrate_for_qwen2.5_vl_7b and difficulty_level_for_qwen2.5_vl_7b. Further details are available in our paper The Synergy Dilemma of Long-CoT SFT and RL: Investigating Post-Training Techniques for Reasoning VLMs.
🚀 Data Usage
from datasets import load_dataset
dataset = load_dataset("JierunChen/MathVista_with_difficulty_level")… See the full description on the dataset page: https://huggingface.co/datasets/JierunChen/MathVista_with_difficulty_level.cleaned-plotqa-v2-difficulty
Cleaned-PlotQA v2 with difficulty tiers (vectorized + calibrated)
This repository augments jrc/cleaned-plotqa-v2 by adding a single column difficulty_tier ∈ {easy, medium, hard} computed with a vectorized, batch‑scored rule set and cutoffs calibrated on a 1,000‑example sample to avoid tier collapse.
Tier counts
easy: 77403
medium: 78521
hard: 43369
total labeled: 199293
Notes
Only one new column is added; original fields remain unchanged.
The scoring runs… See the full description on the dataset page: https://huggingface.co/datasets/Mohamed-Abbas/cleaned-plotqa-v2-difficulty.R2E-Gym-Lite-with-Difficultylabeled_numina_difficultyWe also include data of labeling difficulty from NUMINA, in the following files: labeled_amc_aime_0_-1.json, labeled_math_0_-1.json, labeled_olympiads_0_-1.json.
human_anatomy_qa_with_difficulty
Truth, Trust, and Trouble (TTT) – Medical Anatomy QA Benchmark
This repository hosts the dataset introduced in the EMNLP Industry Track 2025 paper “Truth, Trust, and Trouble: Medical AI on the Edge.”
The dataset contains 1,077 high-quality, clinically validated True/False anatomy questions, designed to evaluate medical LLMs along three critical axes:
Honesty (factual alignment)
Helpfulness (semantic relevance & completeness)
Harmlessness (safety under clinical constraints)
This… See the full description on the dataset page: https://huggingface.co/datasets/ekplatebiryani/human_anatomy_qa_with_difficulty.Korean-DeepMath-with-Difficulty
Korean DeepMath with Difficulty
This dataset enriches ChuGyouk/Korean-DeepMath with difficulty and topic metadata from zwhe99/DeepMath-103K.
Join procedure
Rows are matched using Korean-DeepMath[extra_info][index] -> original DeepMath row index.
Added fields
original_index
difficulty
topic
Intended use
Prepared for controlled Korean mathematical reasoning SFT experiments, including difficulty-aware sampling such as TDCS.
No Easy /… See the full description on the dataset page: https://huggingface.co/datasets/Seungjun/Korean-DeepMath-with-Difficulty.dataset_with_difficulty