Skip to content

📚 All About Agents

Welcome to our comprehensive resource collection for AI agents. This page curates valuable tools, frameworks, research papers, and learning materials to help you understand and build sophisticated agent systems.

Table of Contents

Resource Categories

  1. Agent Papers
  2. Agent Frameworks
  3. Evaluation
  4. Agent Memory
  5. Blogs

🔬 Agent Papers

Research Papers & Publications

Latest research in agent systems, organized by topic area. Papers are ordered roughly chronologically within each category.

🌐 Web Browsing & Search Agents

Papers on web navigation, browsing automation, and search-based agents.

WebThinker: Empowering Large Reasoning Models with Deep Research Capability - Paper · GitHub

WebDancer: Towards Autonomous Information Seeking Agency - Paper

WebSailor: Navigating Super-human Reasoning for Web Agent - Paper

WebShaper: Agentically Data Synthesizing via Information-Seeking Formalization - Paper

WebWatcher: Breaking New Frontier of Vision-Language Deep Research Agent - Paper

BrowseMaster: Towards Scalable Web Browsing via Tool-Augmented Programmatic Agent Pair - Paper · GitHub

Throttling Web Agents Using Reasoning Gates - Paper

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

WebResearcher: Unleashing unbounded reasoning capability in Long-Horizon Agents - Paper

WebWeaver: Structuring Web-Scale Evidence with Dynamic Outlines for Open-Ended Deep Research - Paper

FocusAgent: Simple Yet Effective Ways of Trimming the Large Context of Web Agents - Paper

BrowserAgent: Building Web Agents with Human-Inspired Web Browsing Actions - Paper

WebRouter: Query-specific Router via Variational Information Bottleneck for Cost-sensitive Web Agent - Paper

R-WoM: Retrieval-augmented World Model For Computer-use Agents - Paper

🔎 Deep Research Agents

Papers on deep research systems for complex, long-horizon information-seeking tasks.

DeepResearcher: Scaling Deep Research via Reinforcement Learning in Real-world Environments - Paper · GitHub

SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesis - Paper · GitHub

Deep Researcher with Test-Time Diffusion: Enhancing research capabilities through diffusion-based test-time adaptation - Paper

ReSum: Unlocking Long-Horizon Search Intelligence via Context Summarization - Paper

MARS: Optimizing Dual-System Deep Research via Multi-Agent Reinforcement Learning - Paper

Beyond Turn Limits: Training Deep Search Agents with Dynamic Context Window - Paper

LLM×MapReduce-V3: Enabling Interactive In-Depth Survey Generation through a MCP-Driven Hierarchically Modular Agent System - Paper

DeepPlanner: Scaling Planning Capability for Deep Research Agents via Advantage Shaping - Paper

🤝 Multi-Agent Systems

Papers on multi-agent collaboration, orchestration, and communication.

Profile-Aware Maneuvering: A Dynamic Multi-Agent System for Robust GAIA Problem Solving by AWorld - Paper

AgentOrchestra: Orchestrating Hierarchical Multi-Agent Intelligence with the Tool-Environment-Agent (TEA) Protocol - Paper · GitHub

TUMIX: Multi-Agent Test-Time Scaling with Tool-Use Mixture - Paper

ARM: Discovering Agentic Reasoning Modules for Generalizable Multi-Agent Systems - Paper

CoMAS: Co-Evolving Multi-Agent Systems via Interaction Rewards - Paper

Opponent Shaping in LLM Agents - Paper

A Multi-Agent Framework for Stateful Inference-Time Search - Paper

MOSAIC: Multi-agent Orchestration for Task-Intelligent Scientific Coding - Paper

MASA: LLM-Driven Multi-Agent Systems for Autoformalization - Paper

StoryBox: Collaborative Multi-Agent Simulation for Hybrid Bottom-Up Long-Form Story Generation Using Large Language Models - Paper

MedAgentAudit: Diagnosing and Quantifying Collaborative Failure Modes in Medical Multi-Agent Systems - Paper

🔧 Tool Use & Integration

Papers on tool discovery, tool learning, and tool-augmented agents.

MCP-Zero: Active Tool Discovery for Autonomous LLM Agents - Paper · GitHub

WALT: Web Agents that Learn Tools - Paper

ThinkBrake: Mitigating Overthinking in Tool Reasoning - Paper

TOUCAN: Synthesizing 1.5M Tool-Agentic Trajectories from Real Environments - Paper

ToolBrain: A Flexible RL Framework for Agentic Tools - Paper

MATRIX: Multimodal Agent Tuning for Robust Tool-Use Reasoning - Paper

Adaptive Tool Generation with Models as Tools and Reinforcement Learning - Paper

A²FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning - Paper

🎯 Agent Training & Reinforcement Learning

Papers on RL-based training methods, optimization, and reward modeling for agents.

The Landscape of Agentic Reinforcement Learning for LLMs: A Survey - Paper

Web-Shepherd: Advancing PRMs for Reinforcing Web Agents - Paper · GitHub

Multi-Agent Tool-Integrated Policy Optimization: Enhancing multi-agent systems through integrated tool usage and policy optimization - Paper

Stratified GRPO: Handling Structural Heterogeneity in Reinforcement Learning of LLM Search Agents - Paper

AgentRL: Scaling Agentic Reinforcement Learning with a Multi-Turn, Multi-Task Framework - Paper

DSPO: Stable and Efficient Policy Optimization for Agentic Search and Reasoning - Paper

Can Tool-Integrated Reinforcement Learning Generalize Across Diverse Domains? - Paper

🏗️ Agent Architecture & Workflow Design

Papers on agent architectures, workflow automation, routing, and system design.

AFlow: Automating Agentic Workflow Generation - Paper

AgentFly: Fine-tuning LLM Agents without Fine-tuning LLMs - Paper

Alita: Generalist Agent Enabling Scalable Agentic Reasoning with Minimal Predefinition and Maximal Self-Evolution - Paper · GitHub

Learning to Route: A Rule-Driven Agent Framework for Hybrid-Source Retrieval-Augmented Generation - Paper

Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models - Paper

Open Agent Specification (Agent Spec): Technical Report - Paper

AgentFlow: In-the-Flow Agentic System Optimization: Effective Planning and Tool Use - Paper · GitHub

CaRT: Teaching LLM Agents to Know When They Know Enough - Paper

AlphaApollo: Orchestrating Foundation Models and Professional Tools into a Self-Evolving System for Deep Agentic Reasoning - Paper

🔄 Self-Evolution & Continual Learning

Papers on self-improving agents, experience-based learning, and continual adaptation.

Long Term Memory: The Foundation of AI Self-Evolution - Paper

Scaling Agents via Continual Pre-training: Enhancing agent capabilities through continuous learning approaches - Paper

Towards General Agentic Intelligence via Environment Scaling: Advancing general AI through scalable environment interactions - Paper

Learning on the Job: An Experience-Driven Self-Evolving Agent for Long-Horizon Tasks - Paper

Agent Learning via Early Experience - Paper

Self-Improving LLM Agents at Test-Time - Paper

Dyna-Mind: Learning to Simulate from Experience for Better AI Agents - Paper

Auto-scaling Continuous Memory for GUI Agent - Paper

🏢 Domain-Specific Agents

Papers on agents specialized for specific domains including science, finance, coding, and more.

SciMaster: Towards General-Purpose Scientific AI Agents, Part I. X-Master as Foundation: Can We Lead on Humanity's Last Exam? - Paper · GitHub

SurveyBench: Can LLM(-Agents) Write Academic Surveys that Align with Reader Needs? - Paper

LLM-Based Data Science Agents: A Survey of Capabilities, Challenges, and Future Directions - Paper

QuantAgents: Towards Multi-agent Financial System via Simulated Trading - Paper · Project

AudioToolAgent: An Agentic Framework for Audio-Language Models - Paper

Customer-R1: Personalized Simulation of Human Behaviors via RL-based LLM Agent in Online Shopping - Paper

AutoMLGen: Navigating Fine-Grained Optimization for Coding Agents - Paper

NavSpace: How Navigation Agents Follow Spatial Intelligence Instructions - Paper

VoiceAgentBench: Are Voice Assistants ready for agentic tasks? - Paper

TinyScientist: An Interactive, Extensible, and Controllable Framework for Building Research Agents - Paper

Democratizing AI Scientists using ToolUniverse - Paper · GitHub

DeepTravel: An End-to-End Agentic Reinforcement Learning Framework for Autonomous Travel Planning Agents - Paper

AGENTIQL: An Agent-Inspired Multi-Expert Framework for Text-to-SQL Generation - Paper

FML-bench: A Benchmark for Automatic ML Research Agents Highlighting the Importance of Exploration Breadth - Paper

📖 Surveys & Overviews

Comprehensive survey papers and literature reviews on agent systems.

Deep Research Agents: A Systematic Examination And Roadmap - Paper · GitHub

Small Language Models for Agentic Systems: A Survey of Architectures, Capabilities, and Deployment Trade-offs - Paper

A Survey on Agentic Security: Applications, Threats and Defenses - Paper

A Survey on Agentic Multimodal Large Language Models - Paper

Haystack Engineering: Context Engineering for Heterogeneous and Agentic Long-Context Evaluation - Paper · GitHub

🛡️ Agent Safety & Security

Papers on adversarial attacks, safety concerns, and security of agent systems.

ToolTweak: An Attack on Tool Selection in LLM-Based Agents - Paper

Exploiting Web Search Tools of AI Agents for Data Exfiltration - Paper

HackWorld: Evaluating Computer-Use Agents on Exploiting Web Application Vulnerabilities - Paper

Deep Research Brings Deeper Harm - Paper

🛠️ Agent Frameworks

Popular Agent Development Frameworks

Comprehensive frameworks for building and deploying AI agents across different domains.

F001 - MiroFlow: Build, manage, and scale your AI agents with ease - GitHub

F002 - Youtu-Agent: A simple yet powerful agent framework that delivers with open-source models - GitHub

F003 - OpenManus: No fortress, purely open ground. OpenManus is Coming - GitHub

F004 - OpenBB Platform: Financial data platform for analysts, quants and AI agents - Project

F005 - TradingAgents: Multi-Agents LLM Financial Trading Framework - Paper · GitHub

F006 - JoyAgent-JDGenie: Technical Report on the GAIA - Paper · GitHub

📊 Evaluation

Benchmarks & Evaluation Frameworks

Comprehensive evaluation tools and benchmarks for measuring agent performance across various tasks.

E001 - LiveMCP-101: Stress Testing and Diagnosing MCP-enabled Agents on Challenging Queries - Paper

E002 - BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agent - Paper

E003 - HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering - Paper

E004 - GAIA: a benchmark for General AI Assistants - Paper · Leaderboard

E005 - xbench: Tracking Agents Productivity Scaling with Profession-Aligned Real-World Evaluations - Paper

E006 - MCP-Universe: Benchmarking Large Language Models with Real-World Model Context Protocol Servers - Paper

E007 - FutureX: An Advanced Live Benchmark for LLM Agents in Future Prediction - Paper

E008 - Terminal-Bench: the benchmark for testing AI agents in real terminal environments - GitHub

E009 - Gaia2 and ARE: Empowering the Community to Evaluate Agents - Blog Post

E010 - GPQA: A Graduate-Level Google-Proof Q&A Benchmark - Paper · GitHub

E011 - WebWalkerQA: WebWalker: Benchmarking LLMs in Web Traversal - Paper · GitHub · Leaderboard

E012 - HLE: Humanity's Last Exam - Paper · Website

E013 - BFCL: Berkeley Function Calling Leaderboard - GitHub · Leaderboard

E014 - When2Call: When (not) to Call Tools - Paper · GitHub

E015 - ToolSandbox: A Stateful, Conversational, Interactive Evaluation Benchmark for LLM Tool Use Capabilities - Paper · GitHub

E016 - ToolBeHonest: A Multi-level Hallucination Diagnostic Benchmark for Tool-Augmented Large Language Models - Paper · GitHub

E017 - SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines - Paper · Website

E018 - Terminal-Bench: A benchmark for testing AI agents in terminal environments - Leaderboard · Website

E019 - τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domains - Paper · GitHub

E020 - τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environment - Paper · GitHub

E021 - Deep Research Bench: Evaluating AI Web Research Agents - Paper · Website

E022 - Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents - Paper

E023 - TRAJECT-Bench: A Trajectory-Aware Benchmark for Evaluating Agentic Tool Use - Paper

E024 - ARC-AGI: The General Intelligence Benchmark - Website

E025 - Demystifying Deep Search: A Holistic Evaluation with Hint-Free Multi-Hop Questions and Factorised Metrics - Paper

E026 - BrowseComp-VL: A Comprehensive Benchmark for Vision-Language Web Browsing - Paper

E027 - ACEBench: Who Wins the Match Point in Tool Usage? - Paper

E028 - Haystack Engineering: Context Engineering for Heterogeneous and Agentic Long-Context Evaluation - Paper · GitHub

E029 - DITING: A Multi-Agent Evaluation Framework for Benchmarking Web Novel Translation - Paper

E030 - When Agents Trade: Live Multi-Market Trading Benchmark for LLM Agents - Paper

E031 - A Comprehensive Survey on Benchmarks and Solutions in Software Engineering of LLM-Empowered Agentic System - Paper

E032 - Holistic Agent Leaderboard: The Missing Infrastructure for AI Agent Evaluation - Paper

🧠 Agent Memory

Memory Systems for Persistent Agent Intelligence

Advanced memory solutions for building agents with long-term context and learning capabilities.

M001 - Mem0: Building Production- Ready AI Agents with Scalable Long-Term Memory - GitHub

M002 - memobase: Profile-Based Long-Term Memory for AI Applications - GitHub

M003 - Memento: Fine-tuning LLM Agents without Fine-tuning LLMs - Paper · GitHub

M004 - MEMTRACK: Evaluating Long-Term Memory and State Tracking in Multi-Platform Dynamic Agent Environments - Paper

M005 - A-MEM: Agentic Memory for LLM Agents - Paper · GitHub

M006 - MemoryOS: Memory OS of AI Agent - Paper · GitHub

M007 - Memory-R1: Enhancing Large Language Model Agents to Manage and Utilize Memories via Reinforcement Learning - Paper

M008 - HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models - Paper · GitHub

M009 - MaxKB: Open-source platform for building enterprise-grade agents - GitHub

M010 - MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent - Paper · Website

M011 - LEGOMem: Modular Procedural Memory for Multi-agent LLM Systems for Workflow Automation - Paper

M012 - Memp: Exploring Agent Procedural Memory - Paper

M013 - MIRIX: Multi-Agent Memory System for LLM-Based Agents - Paper · Website

M014 - A-MemGuard: A Proactive Defense Framework for LLM-Based Agent Memory - Paper

M015 - ToolMem: Enhancing Multimodal Agents with Learnable Tool Capability Memory - Paper

M016 - CAM: A Constructivist View of Agentic Memory for LLM-Based Reading Comprehension - Paper

M017 - Mem-α: Learning Memory Construction via Reinforcement Learning - Paper

M018 - Preference-Aware Memory Update for Long-Term LLM Agents - Paper

Blogs

Blog Posts & Tutorials

Curated collection of blog posts, tutorials, and articles about AI agents from various sources and languages.

General Blogs

  • ChatGPT Agent: Introducing ChatGPT Agent

    • Blog Post · OpenAI's latest agent capabilities and features

  • Tongyi DeepResearch: Deep Research Agent for Complex Tasks

    • Blog Post · Alibaba's advanced research agent system

Chinese Blogs

中文博客与资源

精选的中文AI智能体相关博客文章、教程和资源，帮助中文用户更好地理解和应用智能体技术。

  • 17个主流 Agent 框架快速对比

    • 博客链接 · 知乎专栏文章，对比分析主流智能体框架

  • 通义 DeepResearch

    • Blog Post · 阿里巴巴通义智能体深度研究系统介绍

Documentation Info

Last Updated: February 2026 · Doc Contributor: Team @ MiroMind AI