AI Paper Atlas

Research map / Large language models / Reasoning and chain of thought

Thinking in Flow: A Dissipative Stabilization Operator for Robust Autoregressive Reasoning

Yujie Huang, Wenwu He, Zhuo-Xu Cui

ICML 2026 · Large language models › Reasoning and chain of thought · approach: Long CoT dynamics and efficiency

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Recasts autoregressive chain-of-thought decoding as a perturbed long-horizon dynamical system and proposes an inference-time dissipative stabilization operator to suppress drift from irrelevant cues and late corrections.

Problem

Long-horizon chain-of-thought generation in LLMs is brittle under distribution shift and context interference: irrelevant cues persist, small deviations compound into inference drift, and late-stage corrections can destabilize the reasoning trajectory.

Approach

Models autoregressive decoding as a perturbed dynamical system and adds a dissipative stabilization operator at inference time that damps deviations so that errors do not compound along the reasoning trajectory (per the abstract; details not available).

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add TiF controller to SFTLlama-3.2-3B · GSM8K greedy@1accuracy (%): 44.73 -> 50.49Improved
add TiF controller to SFTLlama-3.2-3B · MATH500 greedy@1accuracy (%): 9.00 -> 11.00Improved
add TiF controller to SFTQwen2.5-3B · GSM8K greedy@1accuracy (%): 63.68 -> 66.34Improved
add TiF controller to SFTQwen3-4B · GSM8K greedy@1accuracy (%): 77.41 -> 77.94No clear effect
add TiF controller to SFTLlama-3.1-8B · GSM8K greedy@1accuracy (%): 63.84 -> 64.29No clear effect
add persistent recurrent state over non-recurrent residual of equal size (ablation)TiF · GSM8K greedy@1 (Llama-3.2-3B)accuracy (%): 48.37 -> 50.49Improved
add entropy-based demand Ht to impact gate (ablation)TiF · GSM8K greedy@1 (Llama-3.2-3B)accuracy (%): 46.85 -> 50.49Improved
add capacity supply Ct to impact gate (ablation)TiF · GSM8K greedy@1 (Llama-3.2-3B)accuracy (%): 47.84 -> 50.49Improved
add direction gate rt (ablation)TiF · GSM8K greedy@1 (Llama-3.2-3B)accuracy (%): 47.92 -> 50.49Improved
use ODE state update instead of GRU (ablation)TiF · GSM8K greedy@1 (Llama-3.2-3B)accuracy (%): 45.67 -> 50.49Improved
add Pause (Prefix) tokensSFT · GSM8K greedy@1 (Llama-3.2-3B)accuracy (%): 44.73 -> 42.84Hurt
add NEFTune embedding noiseSFT · GSM8K greedy@1 (Llama-3.2-3B)accuracy (%): 44.73 -> 45.49No clear effect

Experimental setup

Builds on: chain-of-thought prompting

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy