The paper introduces AnytimeReasoner, a framework that optimizes large language models for anytime reasoning by truncating the reasoning process to fit sampled token budgets and requiring the model to summarize the answer for verification. This approach adds verifiable dense rewards, improving credit assignment in reinforcement learning. The authors also propose Budget Relative Policy Optimization (BRPO), a variance reduction technique that enhances learning robustness and efficiency, and demonstrate that their method outperforms GRPO on mathematical reasoning tasks across various budgets and prior distributions.
arXiv:2610.16593v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) implemented with Group Relative Policy Optimization (GRPO) has driven substantial progress in tra...
arXiv:2606. 01281v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has emerged as a powerful paradigm for enhancing the reasoning capabilities of large language models (LLMs).
arXiv:2603. 25184v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) has become essential for post-training large language models (LLMs) in reasoning tasks.
arXiv:2609.40360v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning capabilities of large language models (LLMs), yet their predictions...
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:2601.03822v2 Announce Type: replace Abstract: Large language models (LLMs) can achieve strong reasoning performance with sufficient computation, but they do not inherently know how much computa...
arXiv:2509. 25004v2 Announce Type: replace Abstract: Online reinforcement learning with verifiable rewards (RLVR) has become an effective paradigm for improving the reasoning abilities of large language models, but most methods still optimize reasoning trajectories over the static problem set, wasting rollout budget on solved or overly difficult problems.
The paper introduces APIVIS, a training-time framework that integrates finite-budget Gumbel search into reinforcement learning with verifiable rewards (RLVR) for mathematical reasoning. APIVIS combines direct and searched responses within each rollout group, uses selective supervision on search-improved tokens, and applies value-guided selection to improve verifier rewards at each searched state. Experiments on standard mathematical reasoning benchmarks and various model scales show significant performance gains over existing search-based methods.
arXiv:2606. 04503v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has greatly advanced large reasoning models (LRMs), but it requires timely training on a huge fully-annotated dataset.
Continuous-Utility Direct Preference Optimization (CU‑DPO) replaces binary preference labels with continuous scores to better capture fine‑grained reasoning quality. The framework aligns language models to multiple prompt‑based cognitive strategies, proving a θ(K log K) sample‑complexity improvement and convergence to an entropy‑regularized utility‑maximizing policy. A two‑stage pipeline—strategy selection via best‑vs‑all comparisons and execution refinement using margin‑stratified pairs—yields significant gains on mathematical reasoning, code generation, and causal reasoning benchmarks.
arXiv:2606. 05464v1 Announce Type: new Abstract: Verifiable reward training has improved mathematical and coding reasoning, but these domains capture only part of step-by-step decision making.
arXiv:2602. 08324v5 Announce Type: replace Abstract: Chain-of-Thought (CoT) reasoning successfully enhances the reasoning capabilities of Large Language Models (LLMs), yet it incurs substantial computational overhead for inference.