Research collection

Agent training and self-evolution

Training agents: environment synthesis, self-evolving and self-improving agent loops, agent RL frameworks, and agent post-training recipes.

Papers: 75
Research labs: 6
Official code: 56

51–75 of 75 papers in this collection

51

Meta AI

What Does It Take to Be a Good AI Research Agent? Studying the Role of Ideation Diversity

This paper investigates the role of ideation diversity in AI research agents' performance on MLE-bench, a benchmark of 75 Kaggle machine learning tasks. The authors analyzed 11,000 agent trajectories from 6 LLM backbones and 3 scaffolds, measuring ideation diversity via Shannon entropy of model architectures in initial drafts. They found that…

Alexis Audran-Reiss, Jordi Armengol-Estapé, Karen Hambardzumyan, Amar Budhiraja, et al.

Published: Nov 2025
Citations: 5
Code: Not linked

Read summary Original

52

arXiv.org

Lumine: An Open Recipe for Building Generalist Agents in 3D Open Worlds

Lumine, developed by ByteDance Seed, is the first open recipe for building generalist agents that complete hours-long missions in real time within 3D open-world environments. Built on Qwen2-VL-7B-Base, it processes raw pixels at 5 Hz and generates keyboard-mouse actions at 30 Hz using action chunking, with a hybrid thinking mode that invokes reasoning only…

Weihao Tan, Xiangyang Li, Yunhao Fang, Heyuan Yao, et al.

Published: Nov 2025
Citations: 13
Code: Not linked

Read summary Original

53

arXiv.org

Scaling Agent Learning via Experience Synthesis

DreamGym is a unified framework for scaling reinforcement learning (RL) for LLM-based agents by synthesizing diverse experiences, addressing the high costs and limitations of real-environment rollouts. It uses a reasoning-based experience model that operates in an abstract textual state space, generating consistent state transitions and reward signals via…

Zhaorun Chen, Zhuokai Zhao, Kai Zhang, Bo Liu, et al.

Published: Nov 2025
Citations: 24
Code: Not linked

Read summary Original

54

arXiv.org

Game-TARS: Pretrained Foundation Models for Scalable Generalist Multimodal Game Agents

Game-TARS is a generalist game agent from ByteDance Seed that uses a unified, scalable action space based on native keyboard and mouse inputs, enabling large-scale continual pre-training across OS, web, and simulation games. It was pre-trained on over 500B tokens, including game trajectories, cross-domain agent data, and general multimodal data. Key…

Zihao Wang, Xujing Li, Yining Ye, Junjie Fang, et al.

Published: Oct 2025
Citations: 16
Code: Not linked

Read summary Original

55

arXiv.org

ReCode: Unify Plan and Action for Universal Granularity Control

The paper introduces ReCode, a novel paradigm for LLM-based agents that unifies planning and action within a single code representation to achieve universal control of decision granularity. ReCode treats high-level plans as abstract placeholder functions, which are recursively decomposed into finer-grained sub-functions until reaching executable primitive…

Zhaoyang Yu, Jiayi Zhang, Huixue Su, Yufan Zhao, et al.

Published: Oct 2025
Citations: 11
Code: 561 stars

Read summary Original

56

The Web Conference

DeepAgent: A General Reasoning Agent with Scalable Toolsets

DeepAgent is an end-to-end deep reasoning agent that unifies autonomous thinking, tool discovery, and action execution within a single reasoning process, addressing limitations of predefined workflow-based agents. It introduces an autonomous memory folding mechanism that compresses interaction history into structured episodic, working, and tool memories,…

Xiaoxi Li, Wenxiang Jiao, Jiarui Jin, Guanting Dong, et al.

Published: Oct 2025
Citations: 63
Code: 1.1K stars

Read summary Original

57

Annual Meeting of the Association for Computational Linguistics

Human-Agent Collaborative Paper-to-Page Crafting

AutoPage is a multi-agent system that automatically converts academic papers into interactive project webpages. It uses a coarse-to-fine pipeline with three phases: narrative planning, multimodal content generation, and interactive page rendering. Dedicated 'Checker' agents verify each step against the source paper to prevent hallucination, and optional…

Qianli Ma, Siyu Wang, Yilin Chen, Yinhao Tang, et al.

Published: Oct 2025
Citations: 6
Code: 169 stars

Read summary Original

58

arXiv.org

DeepAnalyze: Agentic Large Language Models for Autonomous Data Science

DeepAnalyze-8B is the first agentic large language model designed for autonomous data science, capable of completing the entire pipeline from raw data sources to analyst-grade research reports. It addresses the limitations of previous workflow-based agents by enabling autonomous orchestration and adaptive optimization through a curriculum-based agentic…

Shaolei Zhang, Ju Fan, Meihao Fan, Guoliang Li, et al.

Published: Oct 2025
Citations: 43
Code: 4.4K stars

Read summary Original

59

arXiv.org

Agent Learning via Early Experience

The paper introduces the 'early experience' paradigm for training language agents, which uses the agent's own actions and resulting future states as supervision without external reward signals. This addresses limitations of both supervised fine-tuning (SFT) on expert data, which is costly and generalizes poorly, and reinforcement learning (RL), which is…

Kai Zhang, Xiangchao Chen, Bo Liu, Tianci Xue, et al.

Published: Oct 2025
Citations: 57
Code: 136 stars

Read summary Original

60

arXiv.org

In-the-Flow Agentic System Optimization for Effective Planning and Tool Use

The paper introduces AGENTFLOW, a trainable agentic framework for planning and tool use that coordinates four specialized modules—planner, executor, verifier, and generator—through an evolving memory. Unlike monolithic tool-integrated reasoning models, AGENTFLOW optimizes its planner on-policy within the multi-turn loop. To address long-horizon credit…

Zhuofeng Li, Haoxiang Zhang, Seungju Han, Sheng Liu, et al.

Published: Oct 2025
Citations: 51
Code: 2K stars

Read summary Original

61

arXiv.org

GEM: A Gym for Agentic LLMs

GEM (General Experience Maker) is an open-source environment simulator for agentic LLMs, analogous to OpenAI-Gym for traditional RL. It provides a standardized environment-agent interface with reset() and step(), asynchronous vectorized execution, and modular wrappers. GEM includes diverse tasks (Math, Code, Game, QA, ReasoningGym, Terminal) and tools…

Zichen Liu, Anya Sims, Keyu Duan, Changyu Chen, et al.

Published: Oct 2025
Citations: 11
Code: 504 stars

Read summary Original

62

arXiv.org

LIMI: Less is More for Agency

The paper introduces LIMI (Less Is More for Intelligent Agency), which challenges the assumption that more training data yields better agentic AI. LIMI demonstrates that sophisticated agentic intelligence can emerge from minimal, strategically curated demonstrations. Using only 78 carefully designed training samples focused on vibe coding and research…

Yang Xiao, Mohan Jiang, Jie Sun, Keyu Li, et al.

Published: Sep 2025
Citations: 17
Code: 162 stars

Read summary Original

63

Annual Meeting of the Association for Computational Linguistics

Towards General Agentic Intelligence via Environment Scaling

This paper introduces AgentScaler, a family of models trained to advance general agentic intelligence through systematic environment scaling. The authors propose a two-stage pipeline: first, they automatically construct diverse, fully simulated environments by collecting over 30,000 APIs, organizing them into domains via community detection, and…

Runnan Fang, Shihao Cai, Baixuan Li, Jialong Wu, et al.

Published: Sep 2025
Citations: 48
Code: Not linked

Read summary Original

64

Qwen

Scaling Agents via Continual Pre-training

The paper introduces Agentic Continual Pre-training (Agentic CPT), a new training stage between pre-training and post-training, to build agentic foundation models for deep research agents. The authors argue that post-training on general-purpose models creates optimization conflicts, as models must simultaneously learn agentic behaviors and align to expert…

Liangcai Su, Zhen Zhang, Guangyu Li, Zhuo Chen, et al.

Published: Sep 2025
Citations: 41
Code: 20K stars

Read summary Original

65

arXiv.org

WebSailor-V2: Bridging the Chasm to Proprietary Agents via Synthetic Data and Scalable Reinforcement Learning

WebSailor-V2 is a post-training pipeline for open-source web agents, comprising data construction (SailorFog-QA-V2), SFT, and RL. The dataset uses a dense knowledge graph with cyclic structures and diverse uncertainty types beyond obfuscation. Training employs a dual-environment RL framework: a high-fidelity simulator for rapid iteration and a managed…

Kuan Li, Zhongwang Zhang, Huifeng Yin, Rui Ye, et al.

Published: Sep 2025
Citations: 60
Code: Not linked

Read summary Original

66

arXiv.org

AgentGym-RL: Training LLM Agents for Long-Horizon Decision Making through Multi-Turn Reinforcement Learning

AgentGym-RL is a unified, modular reinforcement learning framework for training LLM agents in multi-turn, long-horizon decision-making tasks, without requiring supervised fine-tuning. It supports diverse environments (web navigation, deep search, digital games, embodied tasks, scientific tasks) and mainstream RL algorithms (PPO, GRPO, REINFORCE++, RLOO).…

Zhiheng Xi, Jixuan Huang, Chenyang Liao, Baodai Huang, et al.

Published: Sep 2025
Citations: 62
Code: 834 stars

Read summary Original

67

arXiv.org

Mini-o3: Scaling Up Reasoning Patterns and Interaction Turns for Visual Search

Mini-o3 is a system for visual search that scales up tool-based interactions and reasoning patterns, achieving state-of-the-art performance on challenging tasks. It addresses limitations of existing open-source VLMs, which show monotonous reasoning and limited interaction turns. The approach includes three key components: constructing the Visual Probe…

Xin Lai, Junyi Li, Wei Li, Tao Liu, et al.

Published: Sep 2025
Citations: 101
Code: 424 stars

Read summary Original

68

arXiv.org

WebExplorer: Explore and Evolve for Training Long-Horizon Web Agents

WebExplorer introduces a data synthesis framework for training long-horizon web agents. It uses model-based exploration, where LLMs iteratively search and browse from a seed entity to construct an information space, and iterative long-to-short query evolution, which removes salient clues and adds obfuscation to increase query difficulty. This produces the…

Junteng Liu, Yunji Li, Chi Zhang, Jingyang Li, et al.

Published: Sep 2025
Citations: 74
Code: 120 stars

Read summary Original

69

arXiv.org

UI-TARS-2 Technical Report: Advancing GUI Agent with Multi-Turn Reinforcement Learning

UI-TARS-2 is a native GUI-centered agent model developed by ByteDance Seed, designed to handle both structured computer-use tasks and dynamic game environments. It addresses challenges in data scalability, multi-turn reinforcement learning (RL), GUI-only operation limits, and environment stability through a systematic methodology: a data flywheel for…

Haoming Wang, Haoyang Zou, Huatong Song, Jiazhan Feng, et al.

Published: Sep 2025
Citations: 163
Code: 11K stars

Read summary Original

70

arXiv.org

VerlTool: Towards Holistic Agentic Reinforcement Learning with Tool Use

VERLTOOL is a unified, modular framework for Agentic Reinforcement Learning with Tool use (ARLT), addressing fragmentation, synchronous execution bottlenecks, and limited extensibility in existing ARLT codebases. It builds on VeRL, providing upstream alignment, a standardized tool server API supporting diverse tools (code execution, search, SQL, vision,…

Dongfu Jiang, Yi Lu, Zhuofeng Li, Zhiheng Lyu, et al.

Published: Sep 2025
Citations: 81
Code: 1K stars

Read summary Original

71

arXiv.org

AgentScope 1.0: A Developer-Centric Framework for Building Agentic Applications

AgentScope 1.0 is a developer-centric framework for building agentic applications, grounded in the ReAct paradigm. It provides foundational components (message, model, memory, tool) with unified interfaces and extensible modules, supporting multimodal messages, diverse LLM providers, and MCP integration. The framework includes agent-level infrastructure…

Dawei Gao, Zitao Li, Yuexiang Xie, Weirui Kuang, et al.

Published: Aug 2025
Citations: 11
Code: 29K stars

Read summary Original

72

arXiv.org

Mobile-Agent-v3: Fundamental Agents for GUI Automation

The paper introduces GUI-Owl, a foundational GUI agent model built on Qwen2.5-VL, and Mobile-Agent-v3, a multi-agent framework. GUI-Owl-7B achieves state-of-the-art scores among open-source models, including 66.4 on AndroidWorld and 34.9 on OSWorld-Verified, while Mobile-Agent-v3 further improves these to 73.3 and 37.7, respectively. Key innovations…

Jiabo Ye, Xi Zhang, Haiyang Xu, Haowei Liu, et al.

Published: Aug 2025
Citations: 150
Code: 9K stars

Read summary Original

73

arXiv.org

A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems

This survey reviews the emerging paradigm of self-evolving AI agents, which aim to bridge static foundation models with lifelong adaptability. It introduces a unified framework with four components: system inputs, agent system, environment, and optimizers, and proposes 'Three Laws' (Endure, Excel, Evolve) to guide safe and effective evolution. The survey…

Jinyuan Fang, Yanwen Peng, Xi Zhang, Yingxu Wang, et al.

Published: Aug 2025
Citations: 154
Code: 2.4K stars

Read summary Original

74

arXiv.org

Chain-of-Agents: End-to-End Agent Foundation Models via Multi-Agent Distillation and Agentic RL

The paper introduces Chain-of-Agents (CoA), a paradigm enabling a single LLM to perform multi-agent problem-solving end-to-end by dynamically activating role-playing and tool agents. To train these Agent Foundation Models (AFMs), the authors propose multi-agent distillation, which converts trajectories from state-of-the-art multi-agent systems (e.g.,…

Weizhen Li, Jianbo Lin, Zhuosong Jiang, Jingyi Cao, et al.

Published: Aug 2025
Citations: 66
Code: 580 stars

Read summary Original

75

arXiv.org

Agent Lightning: Train ANY AI Agents with Reinforcement Learning

Agent Lightning is a framework that enables reinforcement learning (RL) training of large language models (LLMs) for any AI agent, achieving complete decoupling between agent execution and training with almost zero code modifications. It formulates agent execution as a Markov decision process (MDP), defining a unified data interface that abstracts agent…

Xufang Luo, Yuge Zhang, Zhiyuan He, Zilong Wang, et al.

Published: Aug 2025
Citations: 58
Code: 17K stars

Read summary Original