SyncAI.news, a Varaisys broadcasting
Not All Thinking is Created Equal: Latent Reasoning Discovers a Recurrent Search Algorithm for Depth Generalization
HC

Huzi Cheng, Zhewei Zhang

· 1 min read

ResearcharXiv cs.AI

Not All Thinking is Created Equal: Latent Reasoning Discovers a Recurrent Search Algorithm for Depth Generalization

arXiv:2609.35643v1 Announce Type: new Abstract: Large Language Models can perform multi-step reasoning and improve task performance through different forms of intermediate computation, from token-based traces to computation carried out in latent space. However, a question remains open: do these different forms of thinking rely on the same underlying mechanism? To address this, we train and compare five variants of the same GPTNeoX backbone from scratch on an extended multi-hop reasoning task (ProsQA-Ext): a vanilla model, a Chain-of-Thought (CoT) model, a Pause Token model, and two latent-reasoning models that are optimized end-to-end without intermediate reasoning traces. We find that, strong in-distribution (ID) performance does not guarantee depth generalization. Vanilla, CoT, and Pause Token models solve ID problems well, but rely largely on local graph features and generalize poorly to out-of-distribution (OOD) problems with longer hops. In contrast, latent variants generalize better and show internal dynamics consistent with forward reachability propagation on the graph. Causal interventions and circuit analysis localize this computation to a sparse recurrent search circuit in the bottleneck latent model: an attention head retrieves graph relations, an MLP and the residual stream update the reachability state across recurrent steps, while multiple attention heads together then do the candidate matching. Together, these results show that different thinking mechanisms can learn distinct computational solutions, even at similar ID performance. In this setting, latent recurrence supports a reusable forward-search algorithm that generalizes beyond the training depth.

Original source

This story was published by arXiv cs.AI and written by Huzi Cheng, Zhewei Zhang. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on arxiv.org

Similar News