LoongSage is a lightweight, fully asynchronous reinforcement learning framework designed for large language models that perform agentic tasks. It integrates with Megatron for distributed training and SGLang for high‑throughput inference, centralizing orchestration in a single controller and using Ray actors for GPU components. The framework introduces AgentFlow for unified execution of white‑box and black‑box agents, sliding‑window strategies to limit trajectory staleness, TransferMesh for efficient weight synchronization, and a memory‑efficient Teacher Manager that supports multi‑teacher on‑policy distillation. Experiments across four training runs show steady reward improvement, controlled train‑inference gaps, and sustained hardware utilization by eliminating synchronous idle periods.
arXiv:2608.22167v1 Announce Type: new Abstract: Reinforcement learning (RL) has become an effective way to improve the tool-use ability of large language models (LLMs), but most existing RL framework...
arXiv:2606. 03077v1 Announce Type: cross Abstract: Reinforcement learning (RL) has become a standard post-training paradigm for large language models (LLMs), extending beyond preference alignment to complex reasoning and multi-turn agentic behaviors.
arXiv:2604. 26256v2 Announce Type: replace Abstract: Reinforcement learning (RL) has become a critical paradigm for LLM post-training, yet the rollout phase -- accounting for 50--80% of total step time -- is bottlenecked by skewed generation: long-tailed trajectories indispensable for model performance block the entire training pipeline.
arXiv:2606. 24369v1 Announce Type: new Abstract: Reinforcement learning (RL) has become a dominant post-training paradigm, driving the emergence of high-performance RL systems such as veRL for autoregressive large language models (LLMs).
ReFlow is a system‑algorithm co‑designed framework for efficient reinforcement learning (RL) post‑training of large language models (LLMs). It introduces replay‑driven decoupling, allowing recent trajectories to be reused multiple times to amortize rollout costs while limiting policy staleness. The architecture features a dual replay cycle, an elastic resource pool, and an adaptive scheduler, achieving up to 2.06× higher training throughput compared to existing synchronous and asynchronous RL systems without sacrificing model performance.
Reinforcement learning (RL) has become a dominant post-training paradigm, driving the emergence of high-performance RL systems such as veRL for autoregressive large language models (LLMs). In parallel, diffusion-oriented RL algorithms, e.
arXiv:2512. 22560v2 Announce Type: replace-cross Abstract: Agentic Reinforcement Learning (RL) trains LLMs through multi-turn interactions with environments, producing workloads that mix compute-bound prefill, bandwidth-bound decoding, CPU-heavy environment execution, and bursty reward evaluation.
AsyncFlow is an asynchronous streaming reinforcement learning framework designed to improve the post‑training phase of large language models. It introduces a distributed data storage and transfer module that enables panoramic data management and fine‑grained scheduling, allowing automated pipeline overlapping and dynamic load balancing. The framework also employs an asynchronous producer‑consumer workflow to reduce computational idleness by deferring parameter updates within staleness thresholds, and it is architecturally decoupled from training and inference engines, providing modular, customizable user interfaces. Experiments show an average throughput improvement of 1.59× over the state‑of‑the‑art baseline.
arXiv:2607. 11185v1 Announce Type: new Abstract: Computer use agents (CUAs) are emerging as a powerful interface for automating complex digital workflows through visual perception and GUI execution.
arXiv:2608. 16798v1 Announce Type: cross Abstract: Agent harnesses have substantially improved performance on long-horizon tasks by coordinating agent interactions with the environment.
RL-VLA$^3$ is a fully asynchronous distributed reinforcement learning framework designed for Vision‑Language‑Action (VLA) model training. It allows fine‑grained asynchronous interaction between simulation, inference, and training via dynamic batching schedulers and flexible environment sharding, addressing the variable, resource‑intensive latencies of physical simulators. Experiments across multiple simulation backends, VLA architectures, and RL algorithms show throughput gains of up to 85.2% over synchronous baselines while preserving sample efficiency, and the system scales from 8 to 256 GPUs.
arXiv:2608. 14635v1 Announce Type: cross Abstract: Large language model (LLM) agents are increasingly trained with reinforcement learning in long-horizon, sandboxed environments.