Recursive Self-Improvement (RSI) is the hottest direction in the AI field. In the past six months, many companies have positioned themselves as RSI companies. I came across an excellent RSI research collection, through which you can learn about RSI history and study the latest technologies and research. Subsequent X posts will also introduce some technical companies in this field. https://academy.dair.ai/papers/collections/self-improving-agents…

Self-Improving Agents | DAIR.AI Academy

Source: https://academy.dair.ai/papers/collections/self-improving-agents

1. The original idea

Recursive self-improvement began as a theory question about when a program may safely rewrite itself. The first paper requires a proof before any change. The second co-evolves the environments with the agents, which is where the open-ended curriculum idea starts.

1. 01
2. 02

2. Self-generated training

The first working self-improvement loops changed the weights. A model generates data, a filter or judge keeps the useful outputs, and the model trains on what survived. Across these four papers the filter moves from ground-truth answers, to the model’s own judgment, to tasks the model invents, to training instructions the model writes.

1. 03### STaR: Bootstrapping Reasoning With Reasoning Eric Zelikman et al. · 2022 Generate rationales, keep the ones that reach the correct answer, fine-tune on those, and repeat. Every later self-training loop is a variant of this, and its first author went on to write STOP.
2. 04### Self-Rewarding Language Models 2024 The model scores its own responses as a judge, and iterative training on those scores improves both the answers and the scoring. It removes the frozen reward model, so the evaluator improves along with the policy.
3. 05
4. 06### Self-Adapting Language Models 2025 The model writes self-edits, meaning its own finetuning data and update directives, and reinforcement learning rewards each edit by how well the updated model then performs. The update procedure becomes an output of the model.

3. Learning from experience

Deployed agents usually cannot change their weights, so they improve by writing things down. These papers store lessons, skills and playbooks outside the model and read them back later. The code that writes the notes stays fixed, so these systems are self-improving without being recursive.

1. 07
2. 08
3. 09

4. Improving the improver

Here the field becomes recursive. Each paper applies the improvement process to itself, so the component that proposes changes can also be changed. The same idea appears at three levels, first prompts, then a scaffold program, then whole agent designs.

1. 10### Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution Chrisantha Fernando et al. · 2023 Evolves task prompts using mutation prompts, and evolves the mutation prompts too, so the instructions for improving prompts improve alongside them. It is the first system in the language-model era whose improvement operator is itself under optimization.
2. 11### Automated Design of Agentic Systems 2024 A meta agent writes new agent designs as code, evaluates them, and files them in an archive that informs the next design. It moved the search space from prompts to whole agent programs, and the Darwin Gödel Machine later turned that loop on itself.
3. 12
4. 13

5. Self-modifying code agents

Coding agents made self-modification practical, because the agent’s source is a codebase and coding benchmarks can score the result. These five papers form one lineage. It runs from a single agent editing itself, to open-ended archives of variants, to better rules for choosing which variant to extend, to agents that edit how they improve.

1. 14### Darwin Gödel Machine: Open-Ended Evolution of Self-Improving Agents Jenny Zhang et al. · 2025 Keeps an archive of coding agents, has one rewrite its own code, scores the child on benchmarks, and returns it to the archive. It is the reference system for this branch, and the source of the objective-hacking evidence every builder should read.
2. 15### A Self-Improving Coding Agent Maxime Robeyns, Martin Szummer, Laurence Aitchison · 2025 A coding agent with ordinary tools edits its own codebase and improves from 17% to 53% on a random subset of SWE-bench Verified, with no gradient updates. The simplest demonstration that one agent can be both the system under repair and the engineer doing it.
3. 16
4. 17
5. 18### Hyperagents 2026 Merges the task agent and the meta agent into a single editable program, so the procedure that generates improvements can itself be modified. It drops the assumption that a better coding agent is automatically a better self-improver, which only holds when the task is coding.

6. AI improving AI research

The labs reporting on recursive self-improvement in 2026 mean AI systems that improve the process that trains the next AI system. These seven papers are the primary research behind those claims. They cover discovered programs and training algorithms, faster training infrastructure, automated research papers, and benchmarks for AI research ability.

1. 19### Mathematical discoveries from program search with large language models Bernardino Romera-Paredes et al. · 2023 Evolutionary search over programs an LLM writes, filtered by a systematic evaluator, produced new results in extremal combinatorics. It is the direct predecessor of AlphaEvolve and the first case of a language model making a discovery on an established open problem.
2. 20
3. 21### AlphaEvolve: A coding agent for scientific and algorithmic discovery Alexander Novikov et al. · 2025 An evolutionary coding agent that discovered new algorithms and optimized Google’s own infrastructure. It accelerated training of the model that powers it, which makes it the clearest published case of a deployed system improving its own production pipeline.
4. 22
5. 23
6. 24### PostTrainBench: Can LLM Agents Automate LLM Post-Training? Ben Rank et al. · 2026 Hands an agent one base model, one GPU and ten hours, then asks it to post-train the model on its own. It measures AI automating AI training more directly than any other benchmark here, and audits the reward hacking that follows.
7. 25### Recursive self-improvement of AI research agents Dhruv Srikanth et al. · 2026 AIDE2 improved its own code across an eight-day autonomous run, discovering seven successive changes from a new search policy to memory compression. The gains transferred to held-out benchmarks and reward hacking fell from 55% to 32%.

7. Where the loop breaks

Five results set limits that every self-improving system has to respect. A model trained on its own output loses information about the original data, a model can improve itself only as far as it can verify its own work, and a loop left to run can drift somewhere no one asked for.

1. 26### Large Language Models Cannot Self-Correct Reasoning Yet Jie Huang et al. · 2023 Without external feedback, a model asked to correct its own reasoning does not improve and often gets worse. This is the empirical limit that separates a self-improvement loop with a real signal from one talking to itself.
2. 27
3. 28
4. 29
5. 30

8. The field map

The September 2026 survey, read last. Twenty of the thirty papers above appear in its reference list, so it works as a map of what surrounds this collection rather than a substitute for it.

1. 31### The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement Yi Duan et al. · 2026 The field map, published September 2026 by 36 authors. It organizes recursive self-improvement into a six-level autonomy ladder, from in-task improvement up to a system that improves how it improves, and scores where current models fall short. Twenty of the papers above appear in its reference list, which is a useful way to see what this collection adds and what it leaves out.

Beyond the papers

The lab positions, the venue, and the code you can run. None of these is a paper, and each links to the source’s own page.

• Harness Engineering for Self-Improvement↗Lilian Weng’s July 2026 essay surveying this frontier, and the best single piece of writing to read before the papers.
• When AI builds itself (Anthropic)↗Defines recursive self-improvement as a system that autonomously designs its own successor, and reports that more than 80% of code merged at Anthropic in May 2026 was written by Claude.
• Research acceleration: the view inside OpenAI↗Reports 3.1 agent-workdays per human workday in mid-August 2026, alongside the statement that OpenAI does not yet know how to safely reach full recursive self-improvement.
• The Economics of Recursive Self-Improvement↗Models the feedback loops and finds them strengthening but not yet self-sustaining.
• ICLR 2026 Workshop on Recursive Self-Improvement↗The field’s first dedicated venue, with 110 accepted papers.
• Sakana AI RSI Lab↗A frontier lab organizing a research group around this idea, with the prior work it builds on listed.
• Darwin Gödel Machine writeup↗Source of the objective-hacking examples, including faked test logs, which the paper itself does not spell out.
• Darwin Gödel Machine code↗Runnable reference implementation for entry 14.
• HyperAgents code↗Runnable reference implementation for entry 18.
• Huxley-Gödel Machine code↗Runnable reference implementation for entry 16.