COBALT is a contextual bandit learning method that reuses offline multi-turn code generation trajectories to train large language models efficiently. It treats multi-turn code generation as a one-step recoverable Markov decision process, collecting partial trajectories from a reference LLM and training the target model to complete each partial trajectory in a single step. Experiments on LiveCodeBench show that COBALT improves Pass@1 scores for R1-Distill 8B and Qwen3 8B by up to 9.0 and 6.2 points, matches the performance of two online RL baselines, and enhances robustness against misleading execution feedback.
arXiv:2606. 09961v1 Announce Type: cross Abstract: Training large language models (LLMs) as autonomous agents via reinforcement learning (RL) has enabled frontier models to achieve superhuman performance in long-horizon tasks.
Reinforcement learning (RL) has become a key component in modern large language models, yet the rollout stage remains the key bottleneck in RL training pipelines. Although Multi-Token Prediction (MTP) offers a natural solution to accelerate rollouts through speculative decoding, many studies have observed that MTP acceptance rates degrade significantly during RL training, leading to limited speedup performance.
arXiv:2606. 12370v1 Announce Type: new Abstract: Reinforcement learning (RL) has become a key component in modern large language models, yet the rollout stage remains the key bottleneck in RL training pipelines.
The paper introduces Reverse‑Turn Policy Optimization (RTPO), a method that restructures multi‑turn agentic reinforcement learning rollouts into sparse reverse trees and updates policies in temporal reverse order. This approach addresses three key instability sources—context mismatch, weak turn‑level credit assignment, and asynchronous policy drift—by aligning each decision with its downstream continuation. Theoretical analysis shows RTPO eliminates context mismatch and drift, reduces credit bias, and converges to recursive optimality, while experiments demonstrate performance gains of 21.50% over trajectory‑level and 10.76% over turn‑level baselines on multi‑turn agentic RL benchmarks.
arXiv:2607. 23765v1 Announce Type: cross Abstract: Large language models (LLMs) achieve impressive performance across multiple domains, but using the most capable model for every query is prohibitive at scale.
arXiv:2511.21638v3 Announce Type: replace Abstract: Practical LLM agents often operate over multi-turn conversations where success is determined only after the full interaction ends. Most multi-turn...
arXiv:2607. 13988v1 Announce Type: new Abstract: Multi-turn agents solve complex tasks through extended sequences of tool interactions before producing a final answer, making credit assignment a fundamental challenge during post-training.
arXiv:2607. 15610v1 Announce Type: cross Abstract: Reinforcement learning (RL) has become a key approach for training LLM agents, yet popular methods such as GRPO/RLOO rely on multiple independently sampled complete trajectories for advantage estimation.
RollVerify is a lightweight reinforcement learning framework that addresses the trade‑off between efficiency and accuracy in long‑tail rollout settings. It introduces an off‑policy shift metric (OPS) to quantify deviation in partially generated trajectories and uses sequence‑level and token‑level verification to truncate invalid suffixes before training. Experiments on mathematical reasoning tasks show that RollVerify matches on‑policy accuracy while cutting training costs, with preliminary evidence from code‑generation tasks.
arXiv:2607. 07508v1 Announce Type: cross Abstract: Reinforcement learning (RL) is becoming increasingly important for post-training large language models (LLMs).
arXiv:2609.24985v1 Announce Type: new Abstract: Multi-turn tool-use failures can hinge on a single model call, yet reward variation alone does not reveal which call would benefit from training. When...
Reinforcement learning (RL) is becoming increasingly important for post-training large language models (LLMs). Previous RL pipelines for LLMs were mostly synchronous and batch-interleaved, which is inefficient for long-horizon agentic tasks.