Awesome-Self-Evolving-Agents
[Survey] A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems
https://github.com/EvoAgentX/Awesome-Self-Evolving-Agents
Last synced: about 9 hours ago
JSON representation
-
2. Multi-Agent Optimisation
-
2.2 π MAS Optimisation
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [πβ―Paper - Agent)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - Verse/ScoreFlow)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper - ai/gptswarm)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - cn/agents)]
- [π Paper
- [π Paper - cai/rSDE-Bench)]
- [πβ―Paper - PersonalAI/Agent_Foundation_Models)]
- [πβ―Paper - PersonalAI/Agent-KB)]
-
2.1 βοΈ Automatic Multi-Agent Construction
- [π Paper - WISC/MetaAgent/)]
-
-
4. Evaluation
-
4.3 π‘ Safety, Alignment, and Robustness for Lifelong / Self-Evolving Agents
- ![Star History Chart - history.com/#EvoAgentX/Awesome-Self-Evolving-Agents&Date&Date)
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - WISC/AutoDAN-Turbo)]
- [π Paper - WISC/AGrail4Agent)]
- ![Star History Chart - history.com/#EvoAgentX/Awesome-Self-Evolving-Agents&Date&Date)
-
4.1 π Benchmark-Based Evaluation
- [π Paper
- [π Paper
- [π Paper - MAS)]
- [π Paper
- [π Paper - ConvAI/tree/main/api-bank)]
- [π Paper - chen/ToolQA)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper - arena-x/webarena)]
- [π Paper - evals)]
- [π Paper - NLP/WebAgent)]
- [π Paper - bench/SWE-bench)]
- [π Paper
- [π Paper - benchmark)]
- [π Paper
- [π Paper
- [π Paper - GSAI/YuLan-SwarmIntell)]
- [π Paper
- [π Paper - ai/OSWorld)]
- [π Paper - research/android_world)]
-
4.2 βοΈ LLM-Based Evaluation
- [π Paper
- [π Paper
- [π Paper
- [π Paper - NLP-SG/Auto-Arena-LLMs)]
- [π Paper
- [π Paper - ai/agent-as-a-judge)]
-
-
Open-Source Framework
-
1. Single-Agent Optimisation
-
1.1 π€ LLM Behaviour Optimisation
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - GPT)]
- [π Paper - rewarding-lm-pytorch)]
- [π Paper - instruct)]
- [π Paper - trajectory-reasoning)]
- [π Paper
- [π Paper
- [π Paper - nlp/mstar)]
- [π Paper
- [π Paper - Huang/R-Zero)]
- [π Paper - rl/spiral)]
- [π Paper - research/siiRL)]
- [π Paper - Zero)]
- [π Paper - R1)]
- [π Paper
- [π Paper - book/SeRL)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - Reward-V2)]
- [π Paper
- [π Paper
- [π Paper - nlp/tree-of-thought-llm)]
- [π Paper - of-thought-llm)]
- [π Paper - NLP-Group/Deductive-Beam-Search)]
- [π Paper - of-thoughts)]
- [π Paper - of-Thought)]
- [πβ―Paper
- [πβ―Paper
-
1.2 π¬ Prompt Optimisation
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - deepmind/opro)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - group/textgrad)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [πβ―Paper
-
1.3 π§ Memory Optimization
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - cas/StructRAG)]
- [π Paper
- [π Paper
- [π Paper
- [πβ―Paper - on-the-Fly/Memento)]
- [πβ―Paper
- [πβ―Paper - seed/m3-agent)]
-
1.4 π§° Tool Optimization
- [π Paper - CVC/GPT4Tools)]
- [π Paper
- [π Paper - trial-and-error)]
- [π Paper
- [π Paper - GX/ToolEVO)]
- [π Paper - Baichuan-MLSystemLab/BUTTON)]
- [π Paper - agent/MAT-Agent)]
- [π Paper
- [π Paper - STL-Lab/Adapting-While-Learning)]
- [π Paper - RL/ReTool)]
- [π Paper
- [π Paper - N1)]
- [π Paper
- [π Paper - Agents/SPORT-Agents)]
- [π Paper - NLPIR/Tool-Star)]
- [π Paper - NLPIR/ARPO)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - Planner)]
- [π Paper - Zero)]
- [π Paper
- [π Paper
- [π Paper - tool/CLOVA-tool)]
- [π Paper
-
1.5 π§° Unified Optimization
- [π Paper - ICALK/ELL-StuLife)]
- [π Paper
-
-
3. Domain-Specific Optimisation
-
3.1 𧬠Biomedicine
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - imvr/MedAgent-Pro)]
- [π Paper
- [π Paper
- [π Paper - whitelab/chemcrow-public)]
- [π Paper
- [π Paper
- [π Paper
- [π Paper
- [π Paper - Hy/GenoMAS)]
-
3.2 π» Programming
- [π Paper
- [π Paper - refine)]
- [π Paper
- [π Paper - Hands-AI/OpenHands)]
- [π Paper
- [πβ―Paper
- [π Paper
- [π Instructions
- [π Paper
- [π Paper
- [π Paper
- [π Paper
-
3.3 Scientific Research
- [πβ―Paper - lab/PiFlow)]
-
3.4 π°π Financial and Legal Research
- [π Paper - Agent)]
- [π Paper - Foundation/FinRobot)]
- [π Paper - ai/agentUniverse)]
- [π Paper - FinAI/FinCon)]
- [π Paper
- [π Paper
- [π Paper - song/LaWGPT)]
- [π Paper - yuexi/AgentCourt)]
-
3.5 π§© Other Domain-Specific Optimisation
- [π Paper
- [π Paper - ai.org.cn/)]
- [π Paper
-
Programming Languages
Categories
Sub Categories
1.1 π€ LLM Behaviour Optimisation
35
2.2 π MAS Optimisation
28
1.4 π§° Tool Optimization
27
1.2 π¬ Prompt Optimisation
22
4.1 π Benchmark-Based Evaluation
21
3.1 𧬠Biomedicine
16
1.3 π§ Memory Optimization
13
3.2 π» Programming
12
4.3 π‘ Safety, Alignment, and Robustness for Lifelong / Self-Evolving Agents
11
3.4 π°π Financial and Legal Research
8
4.2 βοΈ LLM-Based Evaluation
6
3.5 π§© Other Domain-Specific Optimisation
3
1.5 π§° Unified Optimization
2
3.3 Scientific Research
1
2.1 βοΈ Automatic Multi-Agent Construction
1