Recursive Self-Improvement

Follow this topic Ask about this topic

Recursive self-improvement asks how AI systems can use their own experience to make future versions more capable, efficient, or safe. Approaches include revising agent prompts, tools, memory, and code; evolving training data or models; and using search, world models, or learned evaluators to guide changes. Most demonstrated progress comes from bounded tasks with external or held-out checks, such as coding, reasoning, robotics, and scientific research benchmarks. A central challenge is ensuring that gains persist beyond the feedback used to produce them, without benchmark overfitting, unsafe regressions, or goal drift.

Latest papers

84 papers40 papers in the last four weeks, up 167% on the four weeks before. 0.2% of all new papers.

What is Missing from AI Post-Training AI: An Empirical Analysis
Aug 19, 2026Joy Jia Yin Lim, Xin Huang, Hao Peng +5AI Agent Evaluation LLM Agent Self-Improvement
Tsinghua University · University of Electronic Science and Technology of China · Renmin University of China
The Dynamics of Intelligence Explosions
Aug 14, 2026Toby OrdRecursive Self-Improvement
Oxford Martin AI Governance Initiative, at Oxford University
AQuA: Recursively Self-Improving Quantitative Trading Research Agents
Aug 13, 2026Jiacheng Guo, Suozhi Huang, Yunlong Gao +5Quantitative Finance Automated Algorithm Discovery
Princeton University · Ant Group · Stanford University
Mendel Gödel Machine: Recursive Self-Improving Coding Agents via Comparative Evolution
Aug 7, 2026Changzhi Liu, Yilun Liu, Sikuan Yan +2Coding Agents Genetic Programming
University of Electronic Science and Technology of China · Ludwig Maximilian University of Munich · Munich Center for Machine Learning
PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents
Aug 4, 2026Shuhan Xue, Zixin Ding, Yichen Shen +6LLM Agent Memory Long-Horizon Agent Evaluation
Frontis-MA1: Training an AI4AI Model towards Recursive Self-Improvement in Machine Learning Engineering
Jul 30, 2026Junlin Yang, Che Jiang, Yu Fu +21Evolutionary Optimization Self-Improving Agents
Horizon Research, Frontis.AI
RSIBench-Data: Benchmarking Data-Centric Research for Recursive Self-Improvement
Jul 28, 2026Fanqing Meng, Lingxiao Du, Qiguang Chen +4LLM Agent Evaluation LLM Agent Self-Improvement
1Evolvent AI · 2National University of Singapore
AREX: Towards a Recursively Self-Improving Agent for Deep Research
Jul 23, 2026Shuqi Lu, Chaofan Li, Kun Luo +21Agentic RL Deep Research Agents
Beijing Academy of Artificial Intelligence (BAAI)
Cura 1T: Healthcare Foundation Model via Recursive Self-Improvement
Jul 15, 2026Haolin Chen, Leon Qi, Steve Brown +7Medical VLMs LLM Agent Training
actAVA AI Team
Recursive Self-Improvement in AI: From Bounded Self-Refinement to Autonomous Research Loops
Jul 8, 2026Mingguang Chen, Licheng Wang, Bo QuLLM Self-Refinement Human-in-the-Loop Evaluation
University of California, Riverside (UCR) · AlphaAvatar · Illinois Institute of Technology (IIT)
MetaSkill-Evolve: Recursive Self-Improvement of LLM Agents via Two-Timescale Meta-Skill Evolution
Jul 6, 2026Zefeng Wang, Minxi Yan, Jinhe Bi +3LLM Agent Self-Improvement LLM Agent Skill Learning
LMU Munich · The Chinese University of Hong Kong · MCML +1
Self-Reference in Large Language Models: The Introspection Threshold for Recursive Self-Improvement
Jul 5, 2026Jiang Zhang, Bing Yuan, Qian ZhangLanguage Model Introspection Language Model Self-Improvement
School of Systems Science, Beijing Normal University · Swarma Research
Algorithmic Unverifiability of Safety for Fixed and Recursively Self-Improving Systems
Jun 26, 2026Jose Pascual Gumbau MezquitaFormal Verification Self-Improving Agents
Universitat Jaume I, Castelló de la Plana, Spain
The Red Queen Gödel Machine: Co-Evolving Agents and Their Evaluators
Jun 24, 2026Alex Iacob, Andrej Jovanović, William F. Shen +12Evolutionary Optimization AI Agent Evaluation
University of Cambridge · Flower Labs · NVIDIA +2
Recursive Self-Evolving Agents via Held-Out Selection
Jun 17, 2026Michael Nguyen, Quoc Nguyen, Paul VuongLLM Agent Evaluation LLM Agent Workflow Optimization
School of Information Technology Monash University Malaysia
From AGI to ASI
Jun 10, 2026Tim Genewein, Matija Franklin, Alexander Lerchner +11Self-Improving Agents Recursive Self-Improvement
Google DeepMind · University of Waterloo (work conducted while at Google DeepMind) · Australian National University +1
From 0-to-1 to 1-to-N: Reproducible Engineering Evidence for MetaAI Recursive Self-Design
Jun 8, 2026Dun Li, Jiatao Li, Hongzhi LiLLM Agent Self-Improvement Self-Improving Agents
Department of Computing, The Hong Kong Polytechnic University, Hong Kong SAR, China · Merchant Marine College, Shanghai Maritime University, Shanghai 201306, China · College of Big Data and Artificial Intelligence, Chizhou University, Chizhou 247100, China
A Relative-Computability Theory of Self-Improving Agents
Apr 9, 2026Chien-Ping LuSelf-Improving Agents Recursive Self-Improvement
What Does a System Modify When It Modifies Itself?
Mar 29, 2026Florentin KochRecursive Self-Improvement
École Polytechnique, Palaiseau, France
A Theory of LLM Information Susceptibility
Mar 24, 2026Zhuo-Yang Song, Hua Xing ZhuLarge Language Model-Guided Optimization Language Model Scaling
School of Physics, Peking University, Beijing 100871, China · Center for High Energy Physics, Peking University, Beijing 100871, China
Polaris: A Gödel Agent Framework for Small Language Models through Experience-Abstracted Policy Repair
Mar 24, 2026Aditya Kakade, Vivek Srivastava, Shirish KarandeSmall Language Models Self-Improving Agents
TCS Research, India
Hyperagents
Mar 19, 2026Jenny Zhang, Bingchen Zhao, Wannan Yang +5Self-Improving Agents Recursive Self-Improvement
University of British Columbia · Vector Institute · University of Edinburgh +4
SAHOO: Safeguarded Alignment for High-Order Optimization Objectives in Recursive Self-Improvement
Mar 6, 2026Subramanyam Sahoo, Aman Chadha, Vinija Jain +1LLM Safety Alignment LLM Self-Refinement
MARS 4.0 Fellowship, Cambridge AI Safety Hub(CAISH), University of Cambridge · AWS Generative AI Innovation Center, Amazon Web Services, USA · Stanford University +2
SGM: A Statistical Godel Machine for Risk-Controlled Recursive Self-Modification
Oct 11, 2025Xuening Wu, Shenqin Yin, Yanlan Kang +4Automated Machine Learning Self-Improving Agents
Fudan University · University of California, San Diego · Tongji University

Aug 19, 2026J/K move · Enter open · S save

What is Missing from AI Post-Training AI: An Empirical Analysis

Joy Jia Yin Lim, Xin Huang, Hao Peng, Yaxi Lu+4

Tsinghua University · University of Electronic Science and Technology of China · Renmin University of China

AI Agent Evaluation LLM Agent Self-Improvement LLM Agents LLM Post-Training Recursive Self-Improvement

Open paper AskPDF arXiv

Large language model (LLM) agents can now post-train an LLM end-to-end, raising the prospect of recursive self-improvement (RSI). Yet this progress is measured by aggregate benchmark scores, which cannot tell whether an agent executes a fixed plan well or strategically revises the plan when it fails. We separate these two capabilities: execution-level capability, iterating within an established training strategy, and strategy-level capability, revising that strategy as experimental evidence accumulates. Analyzing 1,338 post-training trajectories of frontier agents, we find that agents reliably execute post-training but lock into a default strategy, which follows the agent rather than the task, and only 2.1% of transitions between adjacent training runs ever change strategy. We then test whether the agent lacks experience, reasoning, or the decision to switch. (1) Experience improves execution but not the strategy. (2) Additional reasoning compute yields front-loaded gains on easier tasks but refines, rather than revises, the committed strategy. (3) Human review before training changes which strategy the agent locks into, not whether it locks in, whereas a single mid-run instruction outperforms the agent's own continuation by up to 17.44 points under the same budget. In conclusion, what the agent lacks is the decision to reopen a committed strategy and try another one. Realizing RSI therefore calls for interaction protocols and training signals that make strategy revision an explicit, rewarded decision.