arXiv cs.CL· Darpan Aswal, Thomas Palmeira Ferraz, Yongxin Zhou, Maxime Peyrard·· 13 小时前AI 评分38
可观测模式不等于解释：潜在推理模型的因果-几何分析
Observable Patterns Are Not Explanations: A Causal-Geometric Analysis of Latent Reasoning Models
AI 导读
对 Coconut 和 CODI 两个潜在推理模型（LRM）的对照实验发现，BFS 式前沿、可解码算术计算等潜在状态模式同样出现在缺少递归或课程设计的对照组中，且并不稳定影响模型行为。
正文 · 原文
Abstract:Latent reasoning models (LRMs) replace explicit chain-of-thought with continuous thoughts. Recent work treats observable latent-state patterns, such as BFS-like frontiers and decodable arithmetic computation, as evidence for internal reasoning mechanisms. Evaluating two LRMs (Coconut and CODI) against controls lacking the proposed recurrence or curriculum, we find these patterns also appear in the controls and do not consistently affect behavior. Causal interventions reveal that latent-thought utilization is not binary but graded, scaling with a thought's causal effect on model behavior. Geometric analyses reveal this effect concentrates in low-dimensional subspaces whose step-wise geometry grows more structured as their behavioral influence increases. Latent thoughts should therefore be treated as hidden computation, not hidden explanation: decodability, attention, or static structure alone cannot establish LRM reasoning mechanisms; that requires matched controls and causal tests.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2606.12689 [cs.CL] |
| (or arXiv:2606.12689v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2606.12689 arXiv-issued DOI via DataCite |
Submission history
From: Darpan Aswal [view email]
[v1]
Wed, 10 Jun 2026 21:23:22 UTC (318 KB)
[v2]
Sat, 24 Oct 2026 15:35:35 UTC (2,735 KB)
来源：arXiv cs.CL · arxiv.org