Shadowgraph
TrendingAbout
Sign in

YsTvT/Awesome-Agentic-RL-Papers

107

42 commits

updated Oct 22, 2025

See the code

README

Awesome Agentic RL Papers

A curated collection of research papers on LLM-based agents and agent training methodologies.

Awesome License: MIT

πŸ“‹ Table of Contents

🎯 Agent Tuning

TitleVenueDateCodeStars
Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM AgentsICLR 2026 under review2025-09--
SimpleTIR: End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated ReasoningarXiv2025-09GitHub
Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Credit AssignmentarXiv2025-05GitHub
SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning TasksarXiv2025-05GitHub
Group-in-Group Policy Optimization for LLM Agent TrainingarXiv2025-05GitHub
Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement LearningCOLM 20252025-05GitHub
Towards Efficient Online Tuning of VLM Agents via Counterfactual Soft Reinforcement LearningICML 20252025-05GitHub
RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement LearningarXiv2025-04GitHub
NAT: Enhancing Agent Tuning with Negative SamplesNAACL 20252025-04GitHub
UI-R1: Enhancing Action Prediction of GUI Agents by Reinforcement LearningarXiv2025-03GitHub
GUI-Xplore: Empowering Generalizable GUI Agents with One ExplorationCVPR 20252025-03GitHub
ATLaS: Agent Tuning via Learning Critical StepsACL 20252025-03--
ReSearch: Learning to Reason with Search for LLMs via Reinforcement LearningarXiv2025-03GitHub
A Review of Prominent Paradigms for LLM-Based Agents: Tool Use, Planning, and Feedback LearningCOLING 20252025-01GitHub
AgentRefine: Enhancing Agent Generalization through Refinement TuningICLR 20252025-01--
Multi-modal Agent Tuning: Building a VLM-Driven Agent for Efficient Tool UsageICLR 20252024-12GitHub
WEBRL: Training LLM Web Agents via Self-Evolving Online Curriculum RLICLR 20252024-11GitHub
Star-Agents: Automatic Data Optimization with LLM Agents for Instruction TuningNeurIPS 20242024-11--
Agent Q: Advanced Reasoning and Learning for Autonomous AI AgentsarXiv2024-08GitHub
ShortcutsBench: A Large Scale Real-world Benchmark for API-based AgentsICLR 20252024-07GitHub
AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge BasesNeurIPS 20242024-07GitHub
DigiRL: Training In-The-Wild Device-Control Agents with Autonomous RLNeurIPS 20242024-06GitHub
Fine-Tuning Large Vision-Language Models as Decision Making Agents via RLNeurIPS 20242024-05GitHub
AgentStudio: A Toolkit for Building General Virtual AgentsICLR 20252024-03GitHub
Cradle: Empowering Foundation Agents Towards General Computer ControlarXiv2024-03GitHub
Agent-FLAN: Designing Data and Methods of Effective Agent TuningACL Findings 20242024-03GitHub
Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based AgentsNeurIPS 20242024-02GitHub
ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RLarXiv2024-02GitHub
AgentTuning: Enabling Generalized Agent Abilities for LLMsACL Findings 20242023-10GitHub
FireAct: Toward Language Agent Fine-tuningarXiv2023-10GitHub
EmbodiedGPT: Vision-Language Pre-Training via Embodied Chain of ThoughtNeurIPS 20232023-05GitHub
Reflexion: Language Agents with Verbal Reinforcement LearningNeurIPS 20232023-03GitHub

🎯 Agent Memory

TitleVenueDateCodeStars
CAM: A Constructivist View of Agentic Memory for LLM-Based Reading ComprehensionarXiv2025-10GitHub
MemGen: Weaving Generative Latent Memory for Self-Evolving AgentsarXiv2025-09GitHub
Mem-Ξ±: Learning Memory Construction via Reinforcement LearningarXiv2025-09GitHub
Seeing, Listening, Remembering, and Reasoning: A Multimodal Agent with Long-Term MemoryarXiv2025-08GitHub
MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory AgentarXiv2025-07GitHub
MIRIX: Multi-Agent Memory System for LLM-Based AgentsarXiv2025-07GitHub
G-Memory: Tracing Hierarchical Memory for Multi-Agent SystemsarXiv2025-06GitHub
MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon AgentsarXiv2025-06GitHub
Memory OS of AI AgentarXiv2025-06GitHub
A-Mem: Agentic Memory for LLM AgentsarXiv2025-02GitHub
InfLLM: Training-Free Long-Context Extrapolation for LLMs with an Efficient Context MemoryNeurIPS 20242024-02GitHub
RAP: Retrieval-Augmented Planning with Contextual Memory for Multimodal LLM AgentsNeurIPS 20242024-02GitHub
MemGPT:Towards LLMsasOperating SystemsarXiv2023-10GitHub
Augmenting Language Models with Long-Term MemoryNeurIPS 20232023-06GitHub
MemoryBank: Enhancing Large Language Models with Long-Term MemoryAAAI 20242023-05GitHub
Towards mental time travel: a hierarchical memory for reinforcement learning agentsNeurIPS 20212021-05--

πŸ”„ Multi-turn

Overview and Experience

TitleVenueDateCodeStars
A Practitioner's Guide to Multi-Turn Agentic Reinforcement LearningarXiv2025-10GitHub
A Simple "Try Again" Can Elicit Multi-Turn LLM ReasoningarXiv2025-07GitHub
Done Is Better Than Perfect: Unlocking Efficient Reasoning by Structured Multi-Turn DecompositionarXiv2025-05--

Environment and Framework

TitleVenueDateCodeStars
RLFR: EXTENDING REINFORCEMENT LEARNING FOR LLMS WITH FLOW ENVIRONMENTarXiv2025-10GitHub
Don't Just Fine-tune the Agent, Tune the EnvironmentarXiv2025-10GitHub
Towards General Agentic Intelligence via Environment ScalingarXiv2025-09GitHub
DeepDive: Advancing Deep Search Agents with Knowledge Graphs and Multi-Turn RLarXiv2025-09GitHub
EvoEmo: Towards Evolved Emotional Policies for LLM Agents in Multi-Turn NegotiationarXiv2025-09--
AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making Through Multi-Turn Reinforcement LearningarXiv2025-09GitHub
MUA-RL: Multi-Turn User-Interacting Agent Reinforcement Learning for Agentic Tool UsearXiv2025-08GitHub
RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement LearningarXiv2025-04GitHub
ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RLarXiv2024-02GitHub

Policy Optimization

TitleVenueDateCodeStars
Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM AgentsICLR 2026 under review2025-09--
Agentic Reinforced Policy Optimization (ARPO)arXiv2025-07GitHub
Group-in-Group Policy Optimization for LLM Agent TrainingNeurIPS 20252025-05GitHub
Direct Multi-Turn Preference Optimization for Language AgentsEMNLP 20242024-06GitHub
INFORMATION GAIN-BASED POLICY OPTIMIZATION: A SIMPLE AND EFFECTIVE APPROACH FOR MULTITURN LLM AGENTSICLR2026 Under Review2024-06--

Credit Assignment & Reward

TitleVenueDateCodeStars
HYBRID REWARD NORMALIZATION FOR PROCESSSUPERVISED NON-VERIFIABLE AGENTIC TASKSarXiv2025-09--
PARL-MT: Learning to Call Functions in Multi-Turn Conversation with Progress AwarenessarXiv2025-09--
SPA-RL: Reinforcing LLM Agents via Stepwise Progress AttributionarXiv2025-09[GitHub]
Multi-Turn Code Generation Through Single-Step RewardsarXiv2025-09[GitHub]
Stabilizing Long-term Multi-turn Reinforcement Learning with Gated RewardsarXiv2025-08--
Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Credit AssignmentarXiv2025-05GitHub
Multi-Turn Reinforcement Learning from Preference Human FeedbackNeurIPS 20242024-05--

Specific Domain

TitleVenueDateCodeStars
Scaling up Multi-Turn Off-Policy RL and Multi-Agent Tree Search for LLM Step-ProversarXiv2025-09--
Training Long-Context, Multi-Turn Software Engineering Agents with Reinforcement LearningarXiv2025-08--
Kevin: Multi-Turn RL for Generating CUDA KernelsarXiv2025-07--
SAMA: Towards Multi-Turn Referential Grounded Video Chat with Large Language ModelsNeurIPS 20252025-05GitHub
WEBAGENT-R1: Training Web Agents via End-to-End Multi-Turn Reinforcement LearningEMNLP 20252025-05GitHub
SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning TasksarXiv2025-03GitHub
Multi-Turn Response Selection with Commonsense-Enhanced Language ModelsarXiv2024-07--

🎯 Tool Calling

TitleVenueDateCodeStars
StepTool: Enhancing Multi-Step Tool Usage in LLMs via Step-Grained Reinforcement LearningCIKM 20252025-08GitHub
Tool-Planner: Task Planning with Clusters across Multiple ToolsICLR 20252025-08GitHub
GTool: Graph Enhanced Tool Planning with Large Language ModelarXiv2025-08--
Agent-as-Tool: A Study on the Hierarchical Decision Making with Reinforcement LearningarXiv2025-07--
AutoTIR: Autonomous Tools Integrated Reasoning via Reinforcement LearningarXiv2025-07GitHub
NaviAgent: Bilevel Planning on Tool Dependency Graphs for Function CallingarXiv2025-06--
TreeRL: LLM Reinforcement Learning with On-Policy Tree SearchACL 20252025-06GitHub
Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-EvolutionarXiv2025-05GitHub
Agent RL Scaling Law: Agent RL with Spontaneous Code Execution for Mathematical Problem SolvingarXiv2025-05GitHub
Nemotron-Research-Tool-N1: Exploring Tool-Using Language Models with Reinforced ReasoningarXiv2025-05GitHub
ToRL: Scaling Tool-Integrated RLarXiv2025-03GitHub
ToolDial: Multi-turn Dialogue Generation Method for Tool-Augmented Language ModelsICLR 20252025-03GitHub
OctoTools: An Agentic Framework with Extensible Tools for Complex ReasoningNAACL 20252025-02GitHub
From Exploration to Mastery: Enabling LLMs to Master Tools via Self-Driven InteractionsICLR 20252025-02GitHub
AGILE: A Novel Reinforcement Learning Framework of LLM AgentsNeurIPS 20242024-11GitHub
Toolken+: Improving LLM Tool Usage with Reranking and a Reject OptionEMNLP 20242024-10--
MindSearch: Mimicking Human Minds Elicits Deep AI SearcherarXiv2024-07GitHub
Executable Code Actions Elicit Better LLM AgentsICML 20242024-05GitHub
Learning to Use Tools via Cooperative and Interactive Agents with Large Language ModelsarXiv2024-03--
AnyTool: Self-Reflective, Hierarchical Agents for Large-Scale API CallsarXiv2024-02GitHub
TOOLVERIFIER: Generalization to New Tools via Self-VerificationEMNLP 20242024-02GitHub
ToolkenGPT: Augmenting Frozen Language Models with Massive Tools via Tool EmbeddingsNeurIPS 20232024-01GitHub
ControlLLM: Augment Language Models with Tools by Searching on GraphsECCV 20242023-10GitHub

🎯 Search Agent

TitleVenueDateCodeStars
Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM AgentsICLR 2026 under review2025-09--
SimpleTIR: End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated ReasoningarXiv2025-09GitHub
Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Credit AssignmentarXiv2025-05GitHub
SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning TasksarXiv2025-05GitHub
Group-in-Group Policy Optimization for LLM Agent TrainingarXiv2025-05GitHub
Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement LearningCOLM 20252025-05GitHub
Towards Efficient Online Tuning of VLM Agents via Counterfactual Soft Reinforcement LearningICML 20252025-05GitHub
RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement LearningarXiv2025-04GitHub
NAT: Enhancing Agent Tuning with Negative SamplesNAACL 20252025-04GitHub
UI-R1: Enhancing Action Prediction of GUI Agents by Reinforcement LearningarXiv2025-03GitHub
GUI-Xplore: Empowering Generalizable GUI Agents with One ExplorationCVPR 20252025-03GitHub
ATLaS: Agent Tuning via Learning Critical StepsACL 20252025-03--
ReSearch: Learning to Reason with Search for LLMs via Reinforcement LearningarXiv2025-03GitHub
A Review of Prominent Paradigms for LLM-Based Agents: Tool Use, Planning, and Feedback LearningCOLING 20252025-01GitHub
AgentRefine: Enhancing Agent Generalization through Refinement TuningICLR 20252025-01--
WebWalker: Benchmarking LLMs in Web TraversalACL 20252025-01--
Multi-modal Agent Tuning: Building a VLM-Driven Agent for Efficient Tool UsageICLR 20252024-12GitHub
WEBRL: Training LLM Web Agents via Self-Evolving Online Curriculum RLICLR 20252024-11GitHub
Star-Agents: Automatic Data Optimization with LLM Agents for Instruction TuningNeurIPS 20242024-11--
Agent Q: Advanced Reasoning and Learning for Autonomous AI AgentsarXiv2024-08GitHub
ShortcutsBench: A Large Scale Real-world Benchmark for API-based AgentsICLR 20252024-07GitHub
AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge BasesNeurIPS 20242024-07GitHub
DigiRL: Training In-The-Wild Device-Control Agents with Autonomous RLNeurIPS 20242024-06GitHub
Fine-Tuning Large Vision-Language Models as Decision Making Agents via RLNeurIPS 20242024-05GitHub
AgentStudio: A Toolkit for Building General Virtual AgentsICLR 20252024-03GitHub
Cradle: Empowering Foundation Agents Towards General Computer ControlarXiv2024-03GitHub
Agent-FLAN: Designing Data and Methods of Effective Agent TuningACL Findings 20242024-03GitHub
Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based AgentsNeurIPS 20242024-02GitHub
ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RLarXiv2024-02GitHub
WebDancer: Towards Autonomous Information Seeking AgencyarXiv2025-05--
WebSailor: Navigating Super-human Reasoning for Web AgentarXiv2025-07--
WebShaper: Agentically Data Synthesizing via Information-Seeking FormalizationarXiv2025-07--
AgentTuning: Enabling Generalized Agent Abilities for LLMsACL Findings 20242023-10GitHub
FireAct: Toward Language Agent Fine-tuningarXiv2023-10GitHub
WebArena: A Realistic Web Environment for Building Autonomous AgentsNeurIPS 20232023-07GitHub
GAIA: a benchmark for General AI AssistantsNeurIPS 20232023-07GitHub
Reflexion: Language Agents with Verbal Reinforcement LearningNeurIPS 20232023-03GitHub
ReAct: Synergizing Reasoning and Acting in Language ModelsICLR 20232022-10GitHub
ToolFormer: Language Models Can Teach Themselves to Use ToolsNeurIPS 20232022-07GitHub
WebGPT: Browser-assisted question-answering with human feedbackarXiv2021-12--
--

🎯 Agent Workflow

TitleVenueDateCodeStars
Agentdistill: training-free agent distillation withgeneralizable mcp boxesarXiv2025-06--
Get Experience from Practice: LLM Agents with Record & ReplayarXiv2025-05--
AFlow: Automating Agentic Workflow GenerationICLR 20252024-10GitHub
Agent Workflow MemoryarXiv2024-09GitHub
Agent KB: Leveraging Cross-Domain Experience for Agentic Problem SolvingICML 20252024-07GitHub

🎯 Agent Evaluation & Benchmark

TitleVenueDateCodeStars
MCP-AgentBench: Evaluating Real-World Language Agent Performance with MCP-Mediated ToolsarXiv2025-09--
GitTaskBench: A Benchmark for Code AgentsarXiv2025-08GitHub
MLE-bench: Evaluating Machine Learning Agents on ML EngineeringICLR 20252024-10GitHub
WorFBench: Benchmarking Agentic Workflow GenerationICLR 20252024-10GitHub
Agent-SafetyBench: Evaluating the Safety of LLM AgentsCoRR 20242024-12GitHub
CharacterEval: A Chinese Benchmark for Role-Playing Conversational Agent EvaluationACL 20242024-01GitHub
AgentBench: Evaluating LLMs as AgentsICLR 20242023-08GitHub

Contributors

YsTvT

13 commits

KeibingYang

10 commits

Rouse77

7 commits

ElevenLiy

6 commits

Similar repos

YoungDubbyDu/Awesome-LLM-Agent-Optimization-Papers

This is the reading list for the survey "A Survey on the Optimization of LLM-based Agents ". We…

246

yh-yao/awesome-rl-ai-agents

A curated list of recent progress and resources on Reinforcement Learning for AI Agents.

7

AGI-Edgerunners/LLM-Agents-Papers

A repo lists papers related to LLM based agent

2,347

js-lee-AI/awesome-llm-agent-papers

A curated, continuously updated reading list of 200+ papers on LLM agents: planning, memory, tool…

84

zjunlp/LLMAgentPapers

Must-read Papers on LLM Agents.

3,121

ANative-Lab/Awesome-Self-Evolving-Agents

πŸ“– A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and…

2,493

WooooDyy/LLM-Agent-Paper-List

The paper list of the 86-page SCIS cover paper "The Rise and Potential of Large Language Model…

8,220

FoundationAgents/awesome-foundation-agents

About Awesome things towards foundation agents. Papers / Repos / Blogs / ...

2,223

YsTvT/Awesome-Agentic-RL-Papers

107

42 commits

updated Oct 22, 2025

See the code

README

Awesome Agentic RL Papers

A curated collection of research papers on LLM-based agents and agent training methodologies.

Awesome License: MIT

πŸ“‹ Table of Contents

🎯 Agent Tuning

TitleVenueDateCodeStars
Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM AgentsICLR 2026 under review2025-09--
SimpleTIR: End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated ReasoningarXiv2025-09GitHub
Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Credit AssignmentarXiv2025-05GitHub
SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning TasksarXiv2025-05GitHub
Group-in-Group Policy Optimization for LLM Agent TrainingarXiv2025-05GitHub
Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement LearningCOLM 20252025-05GitHub
Towards Efficient Online Tuning of VLM Agents via Counterfactual Soft Reinforcement LearningICML 20252025-05GitHub
RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement LearningarXiv2025-04GitHub
NAT: Enhancing Agent Tuning with Negative SamplesNAACL 20252025-04GitHub
UI-R1: Enhancing Action Prediction of GUI Agents by Reinforcement LearningarXiv2025-03GitHub
GUI-Xplore: Empowering Generalizable GUI Agents with One ExplorationCVPR 20252025-03GitHub
ATLaS: Agent Tuning via Learning Critical StepsACL 20252025-03--
ReSearch: Learning to Reason with Search for LLMs via Reinforcement LearningarXiv2025-03GitHub
A Review of Prominent Paradigms for LLM-Based Agents: Tool Use, Planning, and Feedback LearningCOLING 20252025-01GitHub
AgentRefine: Enhancing Agent Generalization through Refinement TuningICLR 20252025-01--
Multi-modal Agent Tuning: Building a VLM-Driven Agent for Efficient Tool UsageICLR 20252024-12GitHub
WEBRL: Training LLM Web Agents via Self-Evolving Online Curriculum RLICLR 20252024-11GitHub
Star-Agents: Automatic Data Optimization with LLM Agents for Instruction TuningNeurIPS 20242024-11--
Agent Q: Advanced Reasoning and Learning for Autonomous AI AgentsarXiv2024-08GitHub
ShortcutsBench: A Large Scale Real-world Benchmark for API-based AgentsICLR 20252024-07GitHub
AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge BasesNeurIPS 20242024-07GitHub
DigiRL: Training In-The-Wild Device-Control Agents with Autonomous RLNeurIPS 20242024-06GitHub
Fine-Tuning Large Vision-Language Models as Decision Making Agents via RLNeurIPS 20242024-05GitHub
AgentStudio: A Toolkit for Building General Virtual AgentsICLR 20252024-03GitHub
Cradle: Empowering Foundation Agents Towards General Computer ControlarXiv2024-03GitHub
Agent-FLAN: Designing Data and Methods of Effective Agent TuningACL Findings 20242024-03GitHub
Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based AgentsNeurIPS 20242024-02GitHub
ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RLarXiv2024-02GitHub
AgentTuning: Enabling Generalized Agent Abilities for LLMsACL Findings 20242023-10GitHub
FireAct: Toward Language Agent Fine-tuningarXiv2023-10GitHub
EmbodiedGPT: Vision-Language Pre-Training via Embodied Chain of ThoughtNeurIPS 20232023-05GitHub
Reflexion: Language Agents with Verbal Reinforcement LearningNeurIPS 20232023-03GitHub

🎯 Agent Memory

TitleVenueDateCodeStars
CAM: A Constructivist View of Agentic Memory for LLM-Based Reading ComprehensionarXiv2025-10GitHub
MemGen: Weaving Generative Latent Memory for Self-Evolving AgentsarXiv2025-09GitHub
Mem-Ξ±: Learning Memory Construction via Reinforcement LearningarXiv2025-09GitHub
Seeing, Listening, Remembering, and Reasoning: A Multimodal Agent with Long-Term MemoryarXiv2025-08GitHub
MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory AgentarXiv2025-07GitHub
MIRIX: Multi-Agent Memory System for LLM-Based AgentsarXiv2025-07GitHub
G-Memory: Tracing Hierarchical Memory for Multi-Agent SystemsarXiv2025-06GitHub
MEM1: Learning to Synergize Memory and Reasoning for Efficient Long-Horizon AgentsarXiv2025-06GitHub
Memory OS of AI AgentarXiv2025-06GitHub
A-Mem: Agentic Memory for LLM AgentsarXiv2025-02GitHub
InfLLM: Training-Free Long-Context Extrapolation for LLMs with an Efficient Context MemoryNeurIPS 20242024-02GitHub
RAP: Retrieval-Augmented Planning with Contextual Memory for Multimodal LLM AgentsNeurIPS 20242024-02GitHub
MemGPT:Towards LLMsasOperating SystemsarXiv2023-10GitHub
Augmenting Language Models with Long-Term MemoryNeurIPS 20232023-06GitHub
MemoryBank: Enhancing Large Language Models with Long-Term MemoryAAAI 20242023-05GitHub
Towards mental time travel: a hierarchical memory for reinforcement learning agentsNeurIPS 20212021-05--

πŸ”„ Multi-turn

Overview and Experience

TitleVenueDateCodeStars
A Practitioner's Guide to Multi-Turn Agentic Reinforcement LearningarXiv2025-10GitHub
A Simple "Try Again" Can Elicit Multi-Turn LLM ReasoningarXiv2025-07GitHub
Done Is Better Than Perfect: Unlocking Efficient Reasoning by Structured Multi-Turn DecompositionarXiv2025-05--

Environment and Framework

TitleVenueDateCodeStars
RLFR: EXTENDING REINFORCEMENT LEARNING FOR LLMS WITH FLOW ENVIRONMENTarXiv2025-10GitHub
Don't Just Fine-tune the Agent, Tune the EnvironmentarXiv2025-10GitHub
Towards General Agentic Intelligence via Environment ScalingarXiv2025-09GitHub
DeepDive: Advancing Deep Search Agents with Knowledge Graphs and Multi-Turn RLarXiv2025-09GitHub
EvoEmo: Towards Evolved Emotional Policies for LLM Agents in Multi-Turn NegotiationarXiv2025-09--
AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making Through Multi-Turn Reinforcement LearningarXiv2025-09GitHub
MUA-RL: Multi-Turn User-Interacting Agent Reinforcement Learning for Agentic Tool UsearXiv2025-08GitHub
RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement LearningarXiv2025-04GitHub
ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RLarXiv2024-02GitHub

Policy Optimization

TitleVenueDateCodeStars
Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM AgentsICLR 2026 under review2025-09--
Agentic Reinforced Policy Optimization (ARPO)arXiv2025-07GitHub
Group-in-Group Policy Optimization for LLM Agent TrainingNeurIPS 20252025-05GitHub
Direct Multi-Turn Preference Optimization for Language AgentsEMNLP 20242024-06GitHub
INFORMATION GAIN-BASED POLICY OPTIMIZATION: A SIMPLE AND EFFECTIVE APPROACH FOR MULTITURN LLM AGENTSICLR2026 Under Review2024-06--

Credit Assignment & Reward

TitleVenueDateCodeStars
HYBRID REWARD NORMALIZATION FOR PROCESSSUPERVISED NON-VERIFIABLE AGENTIC TASKSarXiv2025-09--
PARL-MT: Learning to Call Functions in Multi-Turn Conversation with Progress AwarenessarXiv2025-09--
SPA-RL: Reinforcing LLM Agents via Stepwise Progress AttributionarXiv2025-09[GitHub]
Multi-Turn Code Generation Through Single-Step RewardsarXiv2025-09[GitHub]
Stabilizing Long-term Multi-turn Reinforcement Learning with Gated RewardsarXiv2025-08--
Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Credit AssignmentarXiv2025-05GitHub
Multi-Turn Reinforcement Learning from Preference Human FeedbackNeurIPS 20242024-05--

Specific Domain

TitleVenueDateCodeStars
Scaling up Multi-Turn Off-Policy RL and Multi-Agent Tree Search for LLM Step-ProversarXiv2025-09--
Training Long-Context, Multi-Turn Software Engineering Agents with Reinforcement LearningarXiv2025-08--
Kevin: Multi-Turn RL for Generating CUDA KernelsarXiv2025-07--
SAMA: Towards Multi-Turn Referential Grounded Video Chat with Large Language ModelsNeurIPS 20252025-05GitHub
WEBAGENT-R1: Training Web Agents via End-to-End Multi-Turn Reinforcement LearningEMNLP 20252025-05GitHub
SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning TasksarXiv2025-03GitHub
Multi-Turn Response Selection with Commonsense-Enhanced Language ModelsarXiv2024-07--

🎯 Tool Calling

TitleVenueDateCodeStars
StepTool: Enhancing Multi-Step Tool Usage in LLMs via Step-Grained Reinforcement LearningCIKM 20252025-08GitHub
Tool-Planner: Task Planning with Clusters across Multiple ToolsICLR 20252025-08GitHub
GTool: Graph Enhanced Tool Planning with Large Language ModelarXiv2025-08--
Agent-as-Tool: A Study on the Hierarchical Decision Making with Reinforcement LearningarXiv2025-07--
AutoTIR: Autonomous Tools Integrated Reasoning via Reinforcement LearningarXiv2025-07GitHub
NaviAgent: Bilevel Planning on Tool Dependency Graphs for Function CallingarXiv2025-06--
TreeRL: LLM Reinforcement Learning with On-Policy Tree SearchACL 20252025-06GitHub
Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-EvolutionarXiv2025-05GitHub
Agent RL Scaling Law: Agent RL with Spontaneous Code Execution for Mathematical Problem SolvingarXiv2025-05GitHub
Nemotron-Research-Tool-N1: Exploring Tool-Using Language Models with Reinforced ReasoningarXiv2025-05GitHub
ToRL: Scaling Tool-Integrated RLarXiv2025-03GitHub
ToolDial: Multi-turn Dialogue Generation Method for Tool-Augmented Language ModelsICLR 20252025-03GitHub
OctoTools: An Agentic Framework with Extensible Tools for Complex ReasoningNAACL 20252025-02GitHub
From Exploration to Mastery: Enabling LLMs to Master Tools via Self-Driven InteractionsICLR 20252025-02GitHub
AGILE: A Novel Reinforcement Learning Framework of LLM AgentsNeurIPS 20242024-11GitHub
Toolken+: Improving LLM Tool Usage with Reranking and a Reject OptionEMNLP 20242024-10--
MindSearch: Mimicking Human Minds Elicits Deep AI SearcherarXiv2024-07GitHub
Executable Code Actions Elicit Better LLM AgentsICML 20242024-05GitHub
Learning to Use Tools via Cooperative and Interactive Agents with Large Language ModelsarXiv2024-03--
AnyTool: Self-Reflective, Hierarchical Agents for Large-Scale API CallsarXiv2024-02GitHub
TOOLVERIFIER: Generalization to New Tools via Self-VerificationEMNLP 20242024-02GitHub
ToolkenGPT: Augmenting Frozen Language Models with Massive Tools via Tool EmbeddingsNeurIPS 20232024-01GitHub
ControlLLM: Augment Language Models with Tools by Searching on GraphsECCV 20242023-10GitHub

🎯 Search Agent

TitleVenueDateCodeStars
Harnessing Uncertainty: Entropy-Modulated Policy Gradients for Long-Horizon LLM AgentsICLR 2026 under review2025-09--
SimpleTIR: End-to-End Reinforcement Learning for Multi-Turn Tool-Integrated ReasoningarXiv2025-09GitHub
Reinforcing Multi-Turn Reasoning in LLM Agents via Turn-Level Credit AssignmentarXiv2025-05GitHub
SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning TasksarXiv2025-05GitHub
Group-in-Group Policy Optimization for LLM Agent TrainingarXiv2025-05GitHub
Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement LearningCOLM 20252025-05GitHub
Towards Efficient Online Tuning of VLM Agents via Counterfactual Soft Reinforcement LearningICML 20252025-05GitHub
RAGEN: Understanding Self-Evolution in LLM Agents via Multi-Turn Reinforcement LearningarXiv2025-04GitHub
NAT: Enhancing Agent Tuning with Negative SamplesNAACL 20252025-04GitHub
UI-R1: Enhancing Action Prediction of GUI Agents by Reinforcement LearningarXiv2025-03GitHub
GUI-Xplore: Empowering Generalizable GUI Agents with One ExplorationCVPR 20252025-03GitHub
ATLaS: Agent Tuning via Learning Critical StepsACL 20252025-03--
ReSearch: Learning to Reason with Search for LLMs via Reinforcement LearningarXiv2025-03GitHub
A Review of Prominent Paradigms for LLM-Based Agents: Tool Use, Planning, and Feedback LearningCOLING 20252025-01GitHub
AgentRefine: Enhancing Agent Generalization through Refinement TuningICLR 20252025-01--
WebWalker: Benchmarking LLMs in Web TraversalACL 20252025-01--
Multi-modal Agent Tuning: Building a VLM-Driven Agent for Efficient Tool UsageICLR 20252024-12GitHub
WEBRL: Training LLM Web Agents via Self-Evolving Online Curriculum RLICLR 20252024-11GitHub
Star-Agents: Automatic Data Optimization with LLM Agents for Instruction TuningNeurIPS 20242024-11--
Agent Q: Advanced Reasoning and Learning for Autonomous AI AgentsarXiv2024-08GitHub
ShortcutsBench: A Large Scale Real-world Benchmark for API-based AgentsICLR 20252024-07GitHub
AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge BasesNeurIPS 20242024-07GitHub
DigiRL: Training In-The-Wild Device-Control Agents with Autonomous RLNeurIPS 20242024-06GitHub
Fine-Tuning Large Vision-Language Models as Decision Making Agents via RLNeurIPS 20242024-05GitHub
AgentStudio: A Toolkit for Building General Virtual AgentsICLR 20252024-03GitHub
Cradle: Empowering Foundation Agents Towards General Computer ControlarXiv2024-03GitHub
Agent-FLAN: Designing Data and Methods of Effective Agent TuningACL Findings 20242024-03GitHub
Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based AgentsNeurIPS 20242024-02GitHub
ArCHer: Training Language Model Agents via Hierarchical Multi-Turn RLarXiv2024-02GitHub
WebDancer: Towards Autonomous Information Seeking AgencyarXiv2025-05--
WebSailor: Navigating Super-human Reasoning for Web AgentarXiv2025-07--
WebShaper: Agentically Data Synthesizing via Information-Seeking FormalizationarXiv2025-07--
AgentTuning: Enabling Generalized Agent Abilities for LLMsACL Findings 20242023-10GitHub
FireAct: Toward Language Agent Fine-tuningarXiv2023-10GitHub
WebArena: A Realistic Web Environment for Building Autonomous AgentsNeurIPS 20232023-07GitHub
GAIA: a benchmark for General AI AssistantsNeurIPS 20232023-07GitHub
Reflexion: Language Agents with Verbal Reinforcement LearningNeurIPS 20232023-03GitHub
ReAct: Synergizing Reasoning and Acting in Language ModelsICLR 20232022-10GitHub
ToolFormer: Language Models Can Teach Themselves to Use ToolsNeurIPS 20232022-07GitHub
WebGPT: Browser-assisted question-answering with human feedbackarXiv2021-12--
--

🎯 Agent Workflow

TitleVenueDateCodeStars
Agentdistill: training-free agent distillation withgeneralizable mcp boxesarXiv2025-06--
Get Experience from Practice: LLM Agents with Record & ReplayarXiv2025-05--
AFlow: Automating Agentic Workflow GenerationICLR 20252024-10GitHub
Agent Workflow MemoryarXiv2024-09GitHub
Agent KB: Leveraging Cross-Domain Experience for Agentic Problem SolvingICML 20252024-07GitHub

🎯 Agent Evaluation & Benchmark

TitleVenueDateCodeStars
MCP-AgentBench: Evaluating Real-World Language Agent Performance with MCP-Mediated ToolsarXiv2025-09--
GitTaskBench: A Benchmark for Code AgentsarXiv2025-08GitHub
MLE-bench: Evaluating Machine Learning Agents on ML EngineeringICLR 20252024-10GitHub
WorFBench: Benchmarking Agentic Workflow GenerationICLR 20252024-10GitHub
Agent-SafetyBench: Evaluating the Safety of LLM AgentsCoRR 20242024-12GitHub
CharacterEval: A Chinese Benchmark for Role-Playing Conversational Agent EvaluationACL 20242024-01GitHub
AgentBench: Evaluating LLMs as AgentsICLR 20242023-08GitHub

Similar repos

YoungDubbyDu/Awesome-LLM-Agent-Optimization-Papers

This is the reading list for the survey "A Survey on the Optimization of LLM-based Agents ". We…

246

yh-yao/awesome-rl-ai-agents

A curated list of recent progress and resources on Reinforcement Learning for AI Agents.

7

AGI-Edgerunners/LLM-Agents-Papers

A repo lists papers related to LLM based agent

2,347

js-lee-AI/awesome-llm-agent-papers

A curated, continuously updated reading list of 200+ papers on LLM agents: planning, memory, tool…

84

zjunlp/LLMAgentPapers

Must-read Papers on LLM Agents.

3,121

ANative-Lab/Awesome-Self-Evolving-Agents

πŸ“– A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and…

2,493

WooooDyy/LLM-Agent-Paper-List

The paper list of the 86-page SCIS cover paper "The Rise and Potential of Large Language Model…

8,220

FoundationAgents/awesome-foundation-agents

About Awesome things towards foundation agents. Papers / Repos / Blogs / ...

2,223

Contributors

YsTvT

13 commits

KeibingYang

10 commits

Rouse77

7 commits

ElevenLiy

6 commits