Self-improving AI / Large Language Models Can Self-Improve
Large Language Models Can Self-Improve
“In this work, we demonstrate that an LLM is also capable of self-improving with only unlabeled datasets.”
Large Language Models Can Self-Improve is one of the projects on the self-improving AI map. Below is what it changes about itself, whether the change lasts, and whether it improves the improver, each with the line from its own source.
Where it sits
Columns: what it changes, from the surface to the deepest change. Rows: whether the change outlasts the run. A number is how many projects sit in that square. See it on the 3D map
- What it changes
- Its model weights“fine-tune the LLM using those self-generated solutions as target outputs”
- Does the change last
- Saved for later runs“then use the same model checkpoint for the evaluation on six Out-Of-Domain (OOD) tasks”
- Does it improve the improver
- improves the task agent; no evidence the improver got better“fine-tune the model on these self-generated reasoning-answer data”
- Who sets the goals
- people set tasks and scoring“our method only leverages the questions in the training set, without using ground truth labels.”
- How it was tested
- tested on held-out tasks“but also enhances out-of-domain (OOD) test tasks (AQUA, StrategyQA, MNLI)”
Date is arXiv v1. PaLM-540B answers unlabeled training questions many times, keeps the reasoning paths that agree with the majority answer, and is fine-tuned on them. The paper describes one round of fine-tuning and does not describe repeating it. The questions come from existing datasets; preliminary studies also have the model write extra questions and its own few-shot prompts.
Does it count as RSI?
It depends on the definition. Under the four used on What counts as RSI, by what its own source says:
- Loose (Eliezer Yudkowsky, Intelligence Explosion Microeconomics (2013)): counts
- Durable (Ren et al., self-improvement survey (2026)): counts
- Strict (Roman Yampolskiy, From Seed AI to Technological Singularity (2015)): does not count
- Autonomous (Anthropic, When AI builds itself): does not count
How it got here
Earlier work in this catalog that Large Language Models Can Self-Improve cites, oldest first. Under each: where that work sits on the map, then where Large Language Models Can Self-Improve sits (changes in bold), and what Large Language Models Can Self-Improve's own paper says about it. Its whole family on the citation map
2022 STaR sets itself apart
Model weights · becomes the next version → saved for later runs · improves the task agent
“Our work is similar to Zelikman et al. (2022) where we both propose to fine-tune a model on self-generated CoT data, but our method does not require ground truth labels and shows stronger empirical results with multi-task generalization.”
“Zelikman et al. (2022) proposes better rationale generation by augmenting ground truth answers as hints when predicted an-swers are incorrect.”
2022 Large Language Models Can Self-Improve
Model weights · saved for later runs · improves the task agent
Cited by Promptbreeder 2023, STOP 2023, Darwin Gödel Machine 2025, Ren, self-improvement 2026