Tagged articles

Agentic RL

30 articles · Page 1 of 1

Machine Learning Algorithms & Natural Language Processing

Oct 17, 2026 · Artificial Intelligence

12K GPU-Hour Reality Check: GRPO Outperforms 10 Self-Distillation Methods in Agentic RL

A large-scale empirical study comparing GRPO with 10 self-distillation methods across 4 tasks and 3 model sizes finds no method consistently beats GRPO; some methods show high final scores but suffer training instability, and compute-equivalent comparisons often reverse apparent gains.

Agentic RLEmpirical StudyGRPO

0 likes · 10 min read

12K GPU-Hour Reality Check: GRPO Outperforms 10 Self-Distillation Methods in Agentic RL

Machine Learning Algorithms & Natural Language Processing

Oct 16, 2026 · Artificial Intelligence

A3 Framework: 14B CLI Agent Beats 235B Model via Structured Action Credit Assignment

Fudan and SJTU researchers open-source A3, a reinforcement learning framework that uses AST-based structured credit assignment to train a 14B CLI agent achieving 24.6% exact match on ShellOps, doubling baselines, and outperforming a 235B model on out-of-distribution tasks with minimal computational overhead.

Agentic RLCLI AgentsNeurIPS 2026

0 likes · 18 min read

A3 Framework: 14B CLI Agent Beats 235B Model via Structured Action Credit Assignment

Machine Learning Algorithms & Natural Language Processing

Oct 12, 2026 · Artificial Intelligence

Agentic RL Infra Survey: 11 Chinese Labs' Approaches to Long-Horizon Credit Assignment

This article surveys Agentic RL infrastructure across 11 leading Chinese AI labs, analyzing their architectural choices (colocated vs. disaggregated), async/partial rollout strategies, staleness control, training-inference mismatch solutions, policy optimization algorithms (GRPO, GSPO, CISPO, etc.), sparse reward credit assignment, reward hacking defenses, and environment/sandbox management, revealing six emerging paradigms and unresolved debates.

Agentic RLGRPO VariantsLong-Horizon Credit Assignment

0 likes · 112 min read

Agentic RL Infra Survey: 11 Chinese Labs' Approaches to Long-Horizon Credit Assignment

Wu Shixiong's Large Model Academy

Sep 28, 2026 · Artificial Intelligence

Agentic RL Reward Design: Rule-Based Verifier with 5-Dim Scoring & 11 Guardrails

The article details a rule-based verifier for Agentic RL post-training in after-sales automation, replacing LLM scoring with structured fact-checking across five weighted dimensions and eleven guardrails to prevent reward hacking, achieving 88% human agreement and boosting task success from 8% to 93%.

Agentic RLGRPOGuardrails

0 likes · 23 min read

Agentic RL Reward Design: Rule-Based Verifier with 5-Dim Scoring & 11 Guardrails

PaperAgent

Sep 24, 2026 · Artificial Intelligence

2026 LLM RL Landscape: From PPO to Agentic RL — Algorithms, Trade-offs & Selection Guide

This article surveys the 2026 reinforcement learning landscape for large language models, detailing foundational algorithms (PPO, DPO, GRPO), advanced GRPO variants (DAPO, GSPO, GMPO, GFPO), emerging Agentic RL methods (ARPO, Tree-GRPO), and a practical scenario-based selection table.

Agentic RLAlgorithm SelectionDPO

0 likes · 12 min read

2026 LLM RL Landscape: From PPO to Agentic RL — Algorithms, Trade-offs & Selection Guide

Machine Heart

Sep 23, 2026 · Backend Development

DeepSeek DSec: Scaling Sandbox Infrastructure to 380K Concurrent Agents

DeepSeek's new paper introduces DSec, a production sandbox infrastructure that manages 300 million daily sandbox instances with 380,000 concurrent at peak, using on-demand image loading, composable layers, memory sharing, and CPU QoS to support large-scale agentic reinforcement learning training.

Agentic RLCPU schedulingContainer Orchestration

0 likes · 19 min read

DeepSeek DSec: Scaling Sandbox Infrastructure to 380K Concurrent Agents

Machine Learning Algorithms & Natural Language Processing

Aug 31, 2026 · Artificial Intelligence

Beyond AutoResearch: Co‑Evolving Strategies and Harnesses with EvoTrainer for Autonomous Training Loops

EvoTrainer introduces a self‑evolving training framework that co‑evolves large‑language‑model policies and their training harness, addressing the static‑harness limitations of traditional autonomous RL and demonstrating consistent gains across math, coding, and software‑engineering tasks.

Agentic RLEvoTrainerLLM

0 likes · 15 min read

Beyond AutoResearch: Co‑Evolving Strategies and Harnesses with EvoTrainer for Autonomous Training Loops

Alibaba Cloud Infrastructure

Aug 27, 2026 · Artificial Intelligence

Cutting 60% of Agentic RL Execution Costs with Alibaba Cloud FC Sandbox

The article explains how Agentic Reinforcement Learning workloads like OSWorld need an execution environment that preserves state across hundreds of actions, and shows that integrating Harbor with Alibaba Cloud Function Compute sandbox meets four strict requirements, enables full‑scale evaluation, and reduces compute‑only costs by about 60%.

Agentic RLCloud ComputingFC Cloud Sandbox

0 likes · 23 min read

Cutting 60% of Agentic RL Execution Costs with Alibaba Cloud FC Sandbox

Machine Learning Algorithms & Natural Language Processing

Aug 25, 2026 · Artificial Intelligence

Understanding Multi‑Round Rollouts, Context Reconstruction, and RL Training in Agentic RL

The article analyzes how Agentic RL decouples internal state, protocol requests, and token sequences, explains the inference pipeline, the challenges of preserving prefix relationships across multi‑round rollouts, and details a gateway‑based data collection and credit‑assignment pipeline for reinforcement‑learning training.

Agentic RLPiRL training

0 likes · 13 min read

Understanding Multi‑Round Rollouts, Context Reconstruction, and RL Training in Agentic RL

Machine Heart

Aug 12, 2026 · Artificial Intelligence

How Libra Allocates Resources for Agentic RL Post‑Training and Boosts Throughput Up to 3×

The paper presents Libra, a resource‑management system for Agentic RL post‑training that jointly optimizes training and rollout GPU allocation using a global planner, heterogeneous inference clusters, a causality‑driven multi‑level feedback queue, and an elastic hybrid pool, achieving up to three‑fold throughput gains and up to 2.5× faster reward convergence.

Agentic RLGPU allocationLarge Language Models

0 likes · 13 min read

How Libra Allocates Resources for Agentic RL Post‑Training and Boosts Throughput Up to 3×

Baobao Algorithm Notes

Aug 4, 2026 · Artificial Intelligence

Agentic RL: Cutting‑Edge Techniques from GLM‑5.2 and Qwen

The article dissects recent Agentic RL breakthroughs—including GLM‑5.2’s shift from GRPO to critic‑based PPO, Qwen’s multi‑dimensional verification system, the generative‑critic GenAC, and the on‑policy skill‑distillation method OPID—showing how each tackles long‑trajectory credit assignment, reward hacking, and scalable evaluation across software‑engineering, front‑end, and real‑world tasks.

Agentic RLGLM-5.2Generative Critic

0 likes · 36 min read

Agentic RL: Cutting‑Edge Techniques from GLM‑5.2 and Qwen

Machine Learning Algorithms & Natural Language Processing

Jul 30, 2026 · Artificial Intelligence

From LLM Rollout to Agentic Rollout: Design Insights and Lessons for an Agentic RL Training Framework

The article analyzes the transition from single‑step LLM rollouts to multi‑step Agentic RL rollouts, compares coupled and decoupled architectures, details the roles of Controller, Runtime Manager, Gateway and LLM Server, and discusses token‑level consistency, trajectory reconstruction, and scalability strategies for a production‑grade training pipeline.

Agentic RLGatewayLLM

0 likes · 25 min read

From LLM Rollout to Agentic Rollout: Design Insights and Lessons for an Agentic RL Training Framework

PaperAgent

Jul 16, 2026 · Artificial Intelligence

Best Practices for Training Long‑Horizon Autonomous Agents

This article surveys recent Agentic RL research, extracts practical design principles, and details concrete implementations such as ToRL, AgentGym‑RL, Agent‑R1, StarPO, and AutoForge, highlighting reward design, environment interfaces, scaling strategies, and stability diagnostics for long‑horizon autonomous agents.

Agentic RLLong-Horizon AgentsScaling

0 likes · 14 min read

Best Practices for Training Long‑Horizon Autonomous Agents

DaTaobao Tech

Jul 13, 2026 · Artificial Intelligence

Agentic RL in Taobao Live: From RLVR to Multi‑Agent Reinforcement Learning

The article details how Taobao Live upgraded its static workflow to a low‑latency Agentic architecture, applied AgentTuning distillation and RLVR to curb hallucinations, and introduced a Multi‑Agent RL framework that separates tool‑calling and reply generation, achieving significant gains in factual correctness, helpfulness, and overall performance.

Agentic RLLLMdigital human

0 likes · 23 min read

Agentic RL in Taobao Live: From RLVR to Multi‑Agent Reinforcement Learning

Machine Heart

Jul 2, 2026 · Artificial Intelligence

How AReaL 2.0 Accelerates Self‑Evolving Agents

AReaL 2.0 introduces an online reinforcement‑learning infrastructure that turns real‑world agent interactions into a learning loop, defining three pillars—trajectory data protocol, data proxy, and evolution control plane—to enable agents to not only execute tasks but continuously improve from their own experience.

AReaLAgentic RLLLM Agents

0 likes · 16 min read

How AReaL 2.0 Accelerates Self‑Evolving Agents

Machine Heart

Jun 17, 2026 · Artificial Intelligence

Why RL‑Trained Agents Still Fail to Reason Actively: The Information Self‑Locking Problem

The paper reveals that outcome‑based reinforcement learning often traps LLM agents in an information self‑locking regime where weak action selection and belief tracking prevent proper credit assignment, and introduces AREW, a lightweight advantage‑reweighting method that restores active reasoning across multiple tasks and models.

AREWAgentic RLLLM Agents

0 likes · 24 min read

Why RL‑Trained Agents Still Fail to Reason Actively: The Information Self‑Locking Problem

PaperAgent

May 21, 2026 · Artificial Intelligence

238 Promising Reinforcement‑Learning Ideas Likely to Earn CCF‑A Papers in 2026

The article compiles 238 cutting‑edge reinforcement‑learning ideas across 21 research directions, highlights recent breakthroughs such as Sutton’s Intentional Updates, and provides brief overviews of representative papers—including knowledge‑graph, Kalman‑filter, agentic, LLM‑driven, and world‑model approaches—along with links to the accompanying source code.

Agentic RLKalman filterLLM

0 likes · 6 min read

238 Promising Reinforcement‑Learning Ideas Likely to Earn CCF‑A Papers in 2026

PaperAgent

Apr 30, 2026 · Artificial Intelligence

Why Reinforcement Learning Is the Future: 2026 Top‑Conference RL Paper Collection

The article highlights the rapid rise of reinforcement learning across major 2026 conferences, curates 181 RL papers from eight top venues, and provides detailed summaries of innovative works such as MSRL and MedVR, offering free access to the papers and code.

Agentic RLReward Modelinglarge models

0 likes · 6 min read

Why Reinforcement Learning Is the Future: 2026 Top‑Conference RL Paper Collection

Huawei Cloud Developer Alliance

Apr 13, 2026 · Artificial Intelligence

How AReaL v1.0 Enables Scalable Agentic RL on Ascend NPU with AWEX Weight Sync

The new AReaL v1.0 release brings full Ascend NPU support, detailed installation guides, and a best‑practice example for training a 30B MoE model across four nodes, while the integrated AWEX weight‑sync mechanism dramatically reduces synchronization time, improving efficiency and stability for large‑scale Agentic RL workloads.

AWEXAgentic RLAscend NPU

0 likes · 12 min read

How AReaL v1.0 Enables Scalable Agentic RL on Ascend NPU with AWEX Weight Sync

Machine Learning Algorithms & Natural Language Processing

Mar 28, 2026 · Artificial Intelligence

A Comprehensive Guide to LLM Post‑Training: From RLHF and GRPO to Agentic RL

This article systematically explains the post‑training pipeline for large language models, covering supervised fine‑tuning, RLHF, PPO, GRPO, RLVR, DPO and emerging Agentic RL, while illustrating each method with analogies, detailed workflows, tables, and recent research findings.

Agentic RLDPOGRPO

0 likes · 24 min read

A Comprehensive Guide to LLM Post‑Training: From RLHF and GRPO to Agentic RL

Alibaba Cloud Infrastructure

Mar 16, 2026 · Artificial Intelligence

Scaling Agentic Reinforcement Learning with a Decoupled T‑Architecture Using Verl and Argo Workflows

Agentic reinforcement learning is evolving from simple text generation to complex, scalable agents, but large‑scale deployment faces challenges like massive parallel rollout scheduling and reproducible environments; this article presents a decoupled T‑architecture that separates high‑level RL logic (Verl) from execution orchestration (Argo Workflows) to address these issues.

Agentic RLArgo WorkflowsScalable Reinforcement Learning

0 likes · 10 min read

Scaling Agentic Reinforcement Learning with a Decoupled T‑Architecture Using Verl and Argo Workflows

Machine Learning Algorithms & Natural Language Processing

Mar 6, 2026 · Artificial Intelligence

Why Reasoning and Tool-Use Clash in Agentic RL—and How DART Solves It

Recent studies reveal that in Agentic RL, jointly training reasoning and tool-use on shared parameters creates a persistent negative interaction, with gradients nearly orthogonal, limiting performance; a disentangled tuning approach (DART) using separate LoRA adapters isolates the two abilities and restores gains across benchmarks.

Agentic RLDartGradient Interference

0 likes · 12 min read

Why Reasoning and Tool-Use Clash in Agentic RL—and How DART Solves It

Baobao Algorithm Notes

Feb 24, 2026 · Artificial Intelligence

The Bitter Lesson of Building Agentic RL in Terminal Environments

This article recounts the challenges of moving from single‑step RL with verifiable rewards to multi‑step agentic reinforcement learning in terminal environments, detailing infrastructure design, asynchronous pipelines, data quality checks, masking strategies, curriculum training, chunk‑based optimization, and practical lessons learned from large‑scale experiments.

Agentic RLAsynchronous TrainingEnvironment Augmentation

0 likes · 33 min read

The Bitter Lesson of Building Agentic RL in Terminal Environments

Old Zhang's AI Learning

Feb 19, 2026 · Artificial Intelligence

Inside GLM-5: Training Techniques, Architecture Innovations, and Benchmark Performance

The article dissects GLM-5’s 744B‑parameter MoE design, 28.5 T token training corpus, novel Muon Split and MLA‑256 optimizations, DSA sparse attention, a fully asynchronous RL pipeline, extensive domestic chip adaptation, and benchmark results that place it on par with Claude Opus 4.5 and ahead of Gemini 3 Pro.

AI architectureAgentic RLDSA

0 likes · 13 min read

Inside GLM-5: Training Techniques, Architecture Innovations, and Benchmark Performance

Baobao Algorithm Notes

Feb 4, 2026 · Artificial Intelligence

Mastering Reinforcement Learning: From Basics to Advanced Agentic RL Techniques

This comprehensive guide walks through reinforcement learning fundamentals, MDP modeling, value functions, Bellman equations, and key algorithms such as Q‑learning, REINFORCE, PPO, DPO, and GRPO, then contrasts LLM‑RL with Agentic‑RL and surveys leading industry frameworks and real‑world applications.

Agentic RLLLMMachine Learning

0 likes · 42 min read

Mastering Reinforcement Learning: From Basics to Advanced Agentic RL Techniques

Alimama Tech

Jan 7, 2026 · Artificial Intelligence

Can Text‑Driven Vibe Coding Tame Complex AI Infra? A Deep Dive into GPU Time‑Sharing for Agentic RL

This article examines the limitations of Vibe Coding for large AI infrastructure, proposes a text‑driven, document‑centric workflow, and presents a time‑multiplexed GPU scheduling solution that dramatically improves rollout throughput and reduces timeouts in large‑scale Agentic RL training.

Agentic RLDesign DocumentsGPU scheduling

0 likes · 21 min read

Can Text‑Driven Vibe Coding Tame Complex AI Infra? A Deep Dive into GPU Time‑Sharing for Agentic RL

Baobao Algorithm Notes

Dec 22, 2025 · Artificial Intelligence

Which Agentic RL Framework Wins? A Deep Dive into AReal, Seer, Slime & verl

This article analyzes the training‑efficiency challenges of multi‑turn agentic reinforcement learning and compares four recent open‑source frameworks—AReal (Ant), Seer (Moonshot), Slime (Zhipu) and verl (Bytedance)—examining their asynchronous inference designs, rollout‑train separation, long‑context handling, off‑policy mitigation, and system‑level optimizations to guide framework selection.

Agentic RLAsynchronous InferenceLarge Language Models

0 likes · 18 min read

Which Agentic RL Framework Wins? A Deep Dive into AReal, Seer, Slime & verl

AntTech

Dec 18, 2025 · Artificial Intelligence

How AEnvironment Powers Scalable Agentic RL with a Unified MCP Protocol

AEnvironment is an open‑source, unified environment platform for Agentic Reinforcement Learning that abstracts all resources as services via the MCP protocol, enabling trillion‑scale model training, rapid app generation, benchmark integration, and seamless deployment through a high‑performance ASandbox runtime.

AEnvironmentAgentic RLEnvironment Platform

0 likes · 11 min read

How AEnvironment Powers Scalable Agentic RL with a Unified MCP Protocol

Xiaolong Cloud Tech Team

Oct 22, 2025 · Artificial Intelligence

A Comprehensive Overview of the Latest Advances in Agentic Reinforcement Learning

This article surveys recent Agentic RL research, detailing token templates, reward designs, training tricks, scaling laws, and multi‑stage frameworks that enable large language models to reason, search, and use tools autonomously.

Agentic RLLLM AgentsMulti-Agent Training

0 likes · 28 min read

A Comprehensive Overview of the Latest Advances in Agentic Reinforcement Learning

Data Party THU

Sep 15, 2025 · Artificial Intelligence

Agentic RL: Transforming LLMs into Autonomous Decision‑Making Agents

This survey formalizes the shift from preference‑based reinforcement fine‑tuning to Agentic Reinforcement Learning, defines Agentic RL via MDP/POMDP abstractions, proposes a dual taxonomy of capabilities and task domains, compiles over 500 recent works, and outlines open challenges for scalable, robust AI agents.

AI agentsAgentic RLLLM

0 likes · 12 min read

Agentic RL: Transforming LLMs into Autonomous Decision‑Making Agents