Lattice Deduction Transformer（LDT）：循环 Transformer 逼近逻辑可靠推理
Lattice Deduction Transformers
AI 导读
研究者提出 Lattice Deduction Transformer（LDT），一种通过格投影在前向传播间逼近逻辑可靠推理的循环 Transformer，训练采用模仿搜索式约束求解器推理的 on-policy 流程，并以基于抽象解释的解候选集近似做监督。
正文 · 原文
Abstract:We introduce the Lattice Deduction Transformer (LDT), a recurrent transformer that approximates logically sound deduction by projecting its latent state through a lattice between forward passes. We train on-policy in a process that mirrors deduction in a search-based constraint solver and supervise training via a domain-agnostic approximation of the set of solution candidates based on abstract interpretation. An $800$K-parameter LDT outperforms prior small recurrent reasoners on Sudoku-Extreme with under $2\%$ of their estimated training FLOPs, solving $99.999\%$ of the $422{,}786$ test puzzles. The model is nearly always sound, in that it gives a wrong answer for $4$ of $1.27$M puzzle evaluations, and otherwise returns a correct answer or abstains. A $1.8$M-parameter LDT variant reaches $99.8\%$ accuracy on Maze-Hard. Our code is available at this http URL.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO) |
| Cite as: | arXiv:2605.08605 [cs.LG] |
| (or arXiv:2605.08605v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.08605 arXiv-issued DOI via DataCite |
Submission history
From: Liam Davis [view email]
[v1]
Sat, 9 May 2026 01:55:45 UTC (123 KB)
[v2]
Tue, 3 Nov 2026 20:31:41 UTC (139 KB)
来源：arXiv cs.AI · arxiv.org