datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
PuzzleWorld
Dataset Card for PuzzleWorld
PuzzleWorld is a benchmark of 667 real-world puzzlehunt–style problems designed to evaluate open-ended, multimodal reasoning capabilities of AI models. Curated from Puzzled Pint’s Creative Commons–licensed archives (2010–2025), each puzzle combines text, visual, and structured inputs with no explicitly stated instructions. Solvers must first infer the hidden problem structure from ambiguous clues and then execute a multi-step, creative reasoning… See the full description on the dataset page: https://huggingface.co/datasets/hzli1202/PuzzleWorld.puzzlescript-gists
PuzzleScript Human-Authored Games (Full Gist Corpus)
35,704 human-authored PuzzleScript games — the
complete source text of each — collected from public GitHub gists.
This is the full corpus: every distinct gist is kept, and each row is tagged
with its deduplication cluster so you can reduce to a unique set with a one-line
filter. The deduplication is reproducible from the shipped dedup_master.json +
dedup_master.py; non-vanilla PuzzleScript-Plus files are excluded (listed in… See the full description on the dataset page: https://huggingface.co/datasets/smearle/puzzlescript-gists.GLM-5.2-Logic-Puzzles
GLM-5.2 · Logical Puzzles
6000x traces distilled from GLM-5.2 on High reasoning
Token Count: 5M~?
Distribution:
Puzzles:
•Tokenization blindless ex: counting the r's in strawberry
•Goal reasoning ex: the car wash test (theres no car wash question exactly just prompts like it so its not just benchmaxxing)
•Reading comprehension traps
•Temporal reasoning
•Many other categories not worth mentioning
Prompts… See the full description on the dataset page: https://huggingface.co/datasets/ianncity/GLM-5.2-Logic-Puzzles.chess_puzzles_10k_in_pgn_san
Lichess Puzzle Database …
Lichess Puzzle Database ▸ Mate‑in‑1/2/3 Subset
A 10k slice of the official Lichess puzzle corpus.
Note: The board is updated to the position that arises after the opponent blunders.
Source
Original dataset: https://huggingface.co/datasets/Lichess/chess-puzzles
What’s Inside
Category
Count
Mate‑in‑1
3333
Mate‑in‑2
3333
Mate‑in‑3
3333
Total
9999
Format Improvements
Field
Old
New… See the full description on the dataset page: https://huggingface.co/datasets/quantum24/chess_puzzles_10k_in_pgn_san.mt_puzzles
Multi-Turn Puzzles: Evaluating Interactive Reasoning and Strategic Dialogue in LLMs
MT-Puzzles is a novel benchmark comprising a suite of multi-turn tasks each designed to test specific reasoning, interactive dialogue, and information-seeking abilities:
Word Guess Guess the secret word in min attempts while environment gives feedback on how close the guess is at each turn.
Movie Recommendation: Probe the user to decode the user preference function for N turns. Pick a movie for the… See the full description on the dataset page: https://huggingface.co/datasets/arianhosseini/mt_puzzles.sliding-puzzle-solver-benchmark
Sliding Puzzle Solver Benchmark — Solution Length & Search Nodes
Optimal (and near-optimal) solution length and search effort for the
LK Forge Sliding Puzzle Solver,
which uses IDA* with a Manhattan-distance + linear-conflict heuristic. The 8-puzzle
(3×3) is solved optimally (weight = 1); the 15-puzzle (4×4) uses a lightly-weighted,
near-optimal search. Boards are generated by a seeded random walk of the blank from the goal.
How the solver works:… See the full description on the dataset page: https://huggingface.co/datasets/LKForge/sliding-puzzle-solver-benchmark.chess_puzzles_1k_in_pgn_san
Lichess Puzzle Database …
Lichess Puzzle Database ▸ Mate‑in‑1/2/3 Subset
A 1k slice of the official Lichess puzzle corpus. 10k soon!
Note: The board is updated to the position that arises after the opponent blunders.
Source
Original dataset: https://huggingface.co/datasets/Lichess/chess-puzzles
What’s Inside
Category
Count
Mate‑in‑1
333
Mate‑in‑2
333
Mate‑in‑3
333
Total
999
Format Improvements
Field
Old
New… See the full description on the dataset page: https://huggingface.co/datasets/quantum24/chess_puzzles_1k_in_pgn_san.puzzle_qamath_logic_puzzles_zebralogic_level_4math_logic_puzzles_zebralogic_level_1math_logic_puzzles_zebralogic_level_3math_logic_puzzles_zebralogic_level_2math_logic_puzzles_zebralogic_level_5math_logic_puzzles_grid_rook_level_3math_logic_puzzles_grid_chip_level_4math_logic_puzzles_grid_rook_level_1math_logic_puzzles_grid_knight_level_1chess_puzzles_datasetmath_logic_puzzles_grid_chip_7B_test_outPuzzleCodeBench
Dataset Card
Overview
This dataset contains the anchor split for public browsing and experimentation.
Transfer Split
transfer.jsonl is the hidden evaluation split. It is used to run generated solver code on unseen instances for benchmark evaluation, and is intentionally not exposed in the public viewer split configuration.
license: cc-by-4.0
puzzlezoo
PuzzleZoo
PuzzleZoo is a collection of three reasoning / planning benchmarks designed to evaluate large language models on multi-step procedural problem-solving — the kind of task where one wrong primitive move silently invalidates the rest of the plan.
It is the official evaluation suite for the paper RePoT: Recoverable Program-of-Thought via Checkpoint Repair (Mazaheri, 2026, arXiv:2605.30052), and is released as a standalone benchmark for the broader community.… See the full description on the dataset page: https://huggingface.co/datasets/parsa-mz/puzzlezoo.math_logic_puzzles_grid_rook_level_4math_logic_puzzles_blocked_grid_level_2math_logic_puzzles_grid_chip_level_2math_logic_puzzles_grid_chip_level_3math_logic_puzzles_grid_rook_level_2math_logic_puzzles_grid_rook_level_5math_logic_puzzles_blocked_grid_level_4math_logic_puzzles_grid_chip_level_1math_logic_puzzles_grid_knight_level_2