Metraipolis › archive › dc_bfa18f8a6502

Text Is the Bottleneck: A Small-Scale Replication of LLM Game Playing in Riichi Mahjong — arXiv:2610.34791

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.34791
License (as archived): CC0-1.0
Archived: 2026-10-27T08:11:39.565Z
SHA-256 of the archived text: edfa5d81a7275c9b85f02e1c8f925056081f5733c145b37724da1b52e08ba21f
Stable URL of this page: https://metraipolis.com/documento/dc_bfa18f8a6502

How to cite

Text Is the Bottleneck: A Small-Scale Replication of LLM Game Playing in Riichi Mahjong — arXiv:2610.34791 — arXiv (en), https://arxiv.org/abs/2610.34791 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bfa18f8a6502 (Metraipolis archive dc_bfa18f8a6502, sha256 edfa5d81a7275c9b85f02e1c8f925056081f5733c145b37724da1b52e08ba21f)

BibTeX

@misc{dcbfa18f8a6502,
  title = {Text Is the Bottleneck: A Small-Scale Replication of LLM Game Playing in Riichi Mahjong — arXiv:2610.34791},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.34791},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bfa18f8a6502; sha256 edfa5d81a7275c9b85f02e1c8f925056081f5733c145b37724da1b52e08ba21f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.34791. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We replicate the riichi-mahjong track of "Can Large Language Models Master Complex Card Games?" ( arXiv:2509.01328 ) under a drastically smaller budget: models of at most 2B parameters, one 16 GB consumer GPU, about five GPU-days. (1) With matched data and labels, text is the bottleneck: on the same 100k-decision training set, a 2.6M-parameter convolutional board encoder agrees with the strong AI Mortal on 80.6% of test decisions, above every text-prompt LLM we train (best 77.8%), and places better at the table (2.85 over 1,000 games vs. ~3.0 for the LLMs; paired difference at n=100 not significant). Scaled to 1.71M decisions, a 4.9M-parameter Transformer reaches 94.7% agreement, beats the 0.8B LLMs on identical seeds (-0.45 ranks, sign test p=0.005), and is statistically equivalent to parity against three Mortal champions (TOST within 0.10 ranks, p=6e-4); parity against a teacher-opponent is roughly the ceiling of pure imitation. (2) Training toward the teacher raises teacher agreement more than a 2.5x parameter increase does (+2.5 vs. +1.5 points); on human-record agreement both effects are +1.0, and neither is visible at the table. (3) Few-point agreement differences do not resolve into playing strength within a family: +2.5 agreement points coincide with a paired placement difference of +0.02 +/- 0.12 ranks (p=1.0), and +2.0 points at 2B with -0.06 (p=0.37) - both below the ~0.25-rank resolution of 100 paired games. We also document a paper-code mismatch in the original winner filter, reconstruct its two private engine patches, and show that a divide-and-conquer pipeline makes template-generated Japanese rationales 93% sentence-exact (43% without facts). Per Tenhou's terms we release no trained weights - only code, configurations, and evaluation protocols.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.34791 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-27T07:56:00.239Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).