arXiv:2608. 15445v1 Announce Type: new Abstract: When a reward is correct on every training example yet consistent with more than one goal, a model can acquire an unintended one, a failure known as goal misgeneralization.
CARE: Causally-Aligned Reasoning Exploration for Medical Large Language Models proposes a new framework to improve medical reasoning in LLMs. It introduces two key conditions—Causal Sufficiency and Proximal Learnability—to curate high-quality training trajectories, using agreement-based self-verification and dynamic entropy bounds. Experiments on medical multimodal and text-only benchmarks show that CARE outperforms competitors, reducing incorrect reasoning and enhancing training stability.
arXiv:2606. 17024v1 Announce Type: new Abstract: Sparse reward reinforcement learning (RL) has become a standard tool for improving LLM reasoning, but its success depends critically on the coverage present in the base model.
The paper introduces Test‑Time Policy Optimization (TTPO), an approach that enables large language models to improve mathematical reasoning without relying on ground‑truth labels. TTPO uses majority‑vote pseudo‑labels and an asymmetric objective: it distills rollouts that agree with the pseudo‑label via On‑Policy Self‑Distillation and penalizes disagreeing rollouts with Grouped Reinforcement Learning. Token‑level selection further refines the process, down‑weighting already‑converged positions during distillation and penalizing only confident errors during RL. Experiments show that TTPO matches label‑supervised OPSD on five competition‑level benchmarks, boosts Qwen3‑1.7B from 38.0 % to 45.2 % in test‑time training, and achieves significant gains without explicit reasoning steps, while also generalizing well across tasks.
arXiv:2606. 11209v1 Announce Type: cross Abstract: Visual question answering increasingly requires multi-step reasoning.
arXiv:2609.37119v1 Announce Type: cross Abstract: Recent approaches to reinforcement learning (RL) post-training for large language models increasingly remove the critic to reduce training instabilit...
arXiv:2606. 05174v1 Announce Type: cross Abstract: Large Language Models (LLMs) have shown strong promise in healthcare applications.
The paper introduces Stepwise Marginal Information Gain (MIG), an intrinsic process reward that evaluates how each reasoning step of a large language model (LLM) or vision-language model (VLM) improves the likelihood of the reference answer. MIG rewards only new likelihood maxima, preventing duplicate credit, and is combined with outcome, format, and self‑distillation objectives to guide training. Experiments on eight benchmarks show that this method outperforms outcome‑only reinforcement learning and improves accuracy by up to 4.8 points over binary‑reward training, including a 12.6‑point gain on MathVerse and a 12.9‑point advantage on vision‑language transfer at 7B parameters.
arXiv:2608.30035v1 Announce Type: new Abstract: Self-evolving reasoning frameworks train a Challenger to generate questions exposing a Solver's weaknesses, creating adaptive curricula without human d...
arXiv:2507.21931v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) often produce plausible but poorly-calibrated answers, limiting their reliability on reasoning-intensive tasks....
The paper introduces Test‑Time Policy Optimization (TTPO), a method that enables large language models to improve mathematical reasoning during inference without relying on ground‑truth labels. TTPO uses majority‑vote pseudo‑labels to guide an asymmetric objective: agreeing rollouts are distilled via On‑Policy Self‑Distillation, while disagreeing rollouts are penalized with Grouped Reinforcement Learning, with token‑level selection refining both branches. Experiments show that TTPO matches label‑supervised OPSD on five benchmarks, boosts Qwen3‑1.7B from 38.0 % to 45.2 % in test‑time training, and achieves significant gains without explicit reasoning steps, demonstrating strong cross‑task generalization.
arXiv:2604. 09482v2 Announce Type: replace Abstract: Reasoning in knowledge-intensive domains remains challenging as intermediate steps are often not locally verifiable: unlike math or code, evaluating step correctness may require synthesizing clues across large external knowledge sources.