arXiv AI October 16, 2026 By Weida Liang, Kenji Kawaguchi

Confidence-Calibrated Inference Expansion for Evaluator-Guided Test-Time Reasoning

The paper introduces a compute‑aware three‑action controller for evaluator‑guided reasoning, deciding whether to trust the current answer, fallback, or expand inference. It scores terminal actions by predicted success and uses a rollout‑based surrogate for expansion, explicitly pricing compute cost. Experiments on width expansion and revision tasks show the controller improves utility over fixed policies while reducing online compute by about 30% and 16%, respectively, and demonstrates that evaluator‑guided reasoning is best viewed as resource allocation under uncertainty.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv Machine Learning

Aug 11

Process Supervision of Confidence Margin for Calibrated LLM Reasoning

arXiv:2604. 23333v2 Announce Type: replace Abstract: Scaling test-time computation with reinforcement learning (RL) has emerged as a reliable path to improve large language models (LLM) reasoning ability.

By Liaoyaqi Wang, Chunsheng Zuo, William Jurayj, Benjamin Van Durme, Anqi Liu

llms reinforcement-learning benchmarks safety

arXiv AI

Oct 1

PACE: Policy-Native Adaptive Decision Timing for Long-Horizon Reasoning

arXiv:2605.09860v5 Announce Type: replace Abstract: Long-horizon reasoning requires deciding not only what actions to take, but how many to execute open-loop before replanning. This number, the execu...

By Chen Li, Zhantao Yang, Fangyi Chen, Han Zhang, Anudeepsekhar Bolimera, Marios Savvides

multimodal

arXiv AI

Jun 6

MPCoT: Reward-Guided Multi-Path Latent Reasoning for Test-Time Scalable Vision-Language-Action

arXiv:2606. 06245v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) policies remain brittle in long-horizon and high-uncertainty control, where one-pass action decoding provides limited inference-time deliberation.

By Boyang Zhang, Lianlei Shan

multimodal

Hugging Face Trending Papers

4d ago

How Much Should a VLA Trust Its Reasoning? Step-Level Guidance as a Measured Action-Value Function

The paper introduces CoC-CFG, a method that applies classifier‑free guidance to the reasoning trace of vision‑language‑action (VLA) policies, weighting each Euler step to modulate the influence of reasoning. By treating the weight selection as a Markov decision process with measured action values, the authors develop Q‑adaptive guidance that conservatively improves the controller. Experiments on the Alpasim driving simulator and the Waymo end‑to‑end benchmark show that Q‑adaptive guidance outperforms fixed weights and online RL baselines, achieving a new state‑of‑the‑art score on the hidden test set.

reinforcement-learning multimodal benchmarks

arXiv AI

Aug 12

UPAIR: Diagnosing Reasoning States via Uncertainty-Progress Alignment for Selective Intervention

arXiv:2607. 17188v2 Announce Type: replace Abstract: While test-time scaling improves the problem-solving ability of large reasoning models (LRMs) through additional inference-time computation, it can also exacerbate overthinking and underthinking, which we formulate as reasoning state--action mismatch.

By Cheng Yan, Zhijun Fan, Guangyang Ye, Fan Xu, Xiang Xia, Yawei Wang, Wuyang Zhang

benchmarks safety

arXiv AI

Jul 7

ASK in the Dark: Uncertainty-Gated LLM Assistance under Partial Observability

arXiv:2607. 02686v1 Announce Type: new Abstract: Reinforcement learning agents operating under partial observability must act on incomplete information, making them natural candidates for guidance from small language models (SLMs) that carry broad reasoning priors.

By Juarez Monteiro, Nathan Gavenski, Guilherme Lima, Francisco Galuppo, Odinaldo Rodrigues, Adriano Veloso

llms agents reinforcement-learning

One email a morning, machine-written

Related stories

Process Supervision of Confidence Margin for Calibrated LLM Reasoning

PACE: Policy-Native Adaptive Decision Timing for Long-Horizon Reasoning

MPCoT: Reward-Guided Multi-Path Latent Reasoning for Test-Time Scalable Vision-Language-Action

How Much Should a VLA Trust Its Reasoning? Step-Level Guidance as a Measured Action-Value Function

UPAIR: Diagnosing Reasoning States via Uncertainty-Progress Alignment for Selective Intervention

ASK in the Dark: Uncertainty-Gated LLM Assistance under Partial Observability