ethanjtang/KinGPT
Text Generation • Updated
text stringlengths 0 81 |
|---|
<|position-start|> |
FEN: r6k/pp2r2p/4Rp1Q/3p4/8/1N1P2b1/PqP3PP/7K w - - 0 25 |
Best move (UCI): e6e7 |
Best move (SAN): Rxe7 |
<|position-end|> |
<|position-start|> |
FEN: r6k/pp2R2p/5p1Q/3p4/8/1N1P2b1/P1P3PP/1q5K w - - 1 26 |
Best move (UCI): b3c1 |
Best move (SAN): Nc1 |
<|position-end|> |
<|position-start|> |
FEN: r6k/pp2R2p/5p1Q/3p4/8/3P2b1/P1P3PP/2q4K w - - 0 27 |
Best move (UCI): h6c1 |
Best move (SAN): Qxc1 |
<|position-end|> |
<|position-start|> |
FEN: 5rk1/1p3ppp/pq1Q1b2/8/8/1P3N2/P4PPP/3R2K1 b - - 3 27 |
Best move (UCI): f8d8 |
Best move (SAN): Rd8 |
<|position-end|> |
<|position-start|> |
FEN: 3Q2k1/1p3ppp/pq3b2/8/8/1P3N2/P4PPP/3R2K1 b - - 0 28 |
Best move (UCI): f6d8 |
Best move (SAN): Bxd8 |
<|position-end|> |
<|position-start|> |
FEN: 8/5R2/1p2P3/p4r2/P6p/1P3Pk1/4K3/8 b - - 2 64 |
Best move (UCI): f5e5 |
Best move (SAN): Re5+ |
<|position-end|> |
<|position-start|> |
FEN: 8/5R2/1p2P3/p3r3/P6p/1P3Pk1/8/5K2 b - - 4 65 |
Best move (UCI): e5e6 |
Best move (SAN): Rxe6 |
<|position-end|> |
<|position-start|> |
FEN: r2qr1k1/b1p2ppp/p5n1/P1p1p3/4P1n1/B2P2Pb/3NBP1P/RN1QR1K1 w - - 0 17 |
Best move (UCI): e2g4 |
Best move (SAN): Bxg4 |
<|position-end|> |
<|position-start|> |
FEN: r2qr1k1/b1p2ppp/p5n1/P1p1p3/4P1b1/B2P2P1/3N1P1P/RN1QR1K1 w - - 0 18 |
Best move (UCI): d1g4 |
Best move (SAN): Qxg4 |
<|position-end|> |
<|position-start|> |
FEN: 6k1/5p1p/4p3/4q3/3n4/2Q3P1/PP1N1P1P/6K1 b - - 3 37 |
Best move (UCI): d4e2 |
Best move (SAN): Ne2+ |
<|position-end|> |
<|position-start|> |
FEN: 6k1/5p1p/4p3/4q3/8/2Q3P1/PP1NnP1P/5K2 b - - 5 38 |
Best move (UCI): e2c3 |
Best move (SAN): Nxc3 |
<|position-end|> |
<|position-start|> |
FEN: 2Q2bk1/5p1p/p5p1/2p3P1/4B3/7P/qPr2P2/2K4R w - - 0 33 |
Best move (UCI): e4c2 |
Best move (SAN): Bxc2 |
<|position-end|> |
<|position-start|> |
FEN: 2Q2bk1/5p1p/p5p1/2p3P1/8/7P/1PB2P2/q1K4R w - - 1 34 |
Best move (UCI): c2b1 |
Best move (SAN): Bb1 |
<|position-end|> |
<|position-start|> |
FEN: r4r2/1p3pkp/p7/3R1p1Q/3P4/8/P1q2P2/3R2K1 w - - 0 26 |
Best move (UCI): d5c5 |
Best move (SAN): Rc5 |
<|position-end|> |
<|position-start|> |
FEN: r4r2/1p3pkp/p7/2R2p1Q/3Pq3/8/P4P2/3R2K1 w - - 2 27 |
Best move (UCI): h5g5 |
Best move (SAN): Qg5+ |
<|position-end|> |
<|position-start|> |
FEN: r4r1k/1p3p1p/p7/2R2pQ1/3Pq3/8/P4P2/3R2K1 w - - 4 28 |
Best move (UCI): g5f6 |
Best move (SAN): Qf6+ |
<|position-end|> |
<|position-start|> |
FEN: 8/8/4k1p1/2KpP2P/5P2/8/8/8 b - - 0 53 |
Best move (UCI): g6h5 |
Best move (SAN): gxh5 |
Training/validation split of puzzles used in GAMBIT: Generalization or Memorization? Brittleness Testing for Chess-Trained Language Models
Parsed in text format from the Lichess Puzzle Database
sampling_log.txt — Displays statistics about samples for each theme (74 unique themes)training-puzzle-positions.txt — Large set of puzzle positions as position + best move pairsvalidation_puzzles_THEME — N=1000 sample of puzzles tagged with THEMEFrom: training-puzzle-positions.txt
<|position-start|>
FEN: r2qr1k1/b1p2ppp/p5n1/P1p1p3/4P1n1/B2P2Pb/3NBP1P/RN1QR1K1 w - - 0 17
Best move (UCI): e2g4
Best move (SAN): Bxg4
<|position-end|>
Each training position is separated into paragraphs via a newline character.
From: validation_puzzles_mateIn1.txt
<|puzzle-start|>
<|position-start|>
FEN: r6r/1ppbk3/1b1p4/pP1PpnN1/P1P5/3P2p1/4N1PP/R2Q1R1K b - - 0 24
Best move (UCI): h8h2
Best move (SAN): Rxh2#
<|position-end|>
<|puzzle-end|>
From: validation_puzzles_mateIn3.txt
<|puzzle-start|>
<|position-start|>
FEN: 2r2rk1/3Q1ppp/pq2p3/1p2N3/1P3P2/P1P5/5nPP/RNB2RK1 b - - 4 18
Best move (UCI): f2h3
Best move (SAN): Nh3+
<|position-end|>
<|position-start|>
FEN: 2r2rk1/3Q1ppp/pq2p3/1p2N3/1P3P2/P1P4n/6PP/RNB2R1K b - - 6 19
Best move (UCI): b6g1
Best move (SAN): Qg1+
<|position-end|>
<|position-start|>
FEN: 2r2rk1/3Q1ppp/p3p3/1p2N3/1P3P2/P1P4n/6PP/RNB3RK b - - 0 20
Best move (UCI): h3f2
Best move (SAN): Nf2#
<|position-end|>
<|puzzle-end|>
@misc{tang2026generalizationmemorizationbrittlenesstesting,
title={Generalization or Memorization? Brittleness Testing for Chess-Trained Language Models},
author={Ethan Tang},
year={2026},
eprint={2605.17565},
archivePrefix={arXiv},
primaryClass={cs.AI},
url={https://arxiv.org/abs/2605.17565},
}