Contributing

Contributions that add a relevant, verifiable paper are welcome.

How to add a paper

Add an entry to the section that matches where the paper enters the loop, matching the existing format:

- [Full Paper Title](https://arxiv.org/abs/XXXX.XXXXX) - First-Author Surname et al. Year

Then open a pull request.

Requirements

• A working arXiv or DOI link is mandatory. Prefer the arXiv abstract page over the PDF.
• One line per paper, and one paper per line.
• Put the paper in the section matching its role in the loop, not the venue it appeared in. A reinforcement-learning paper that folds control into the weights belongs in Trained Loops, not Loop Paradigms, even if the abstract talks about planning.
• Do not add a paper twice. If it spans two sections, pick the one where its main contribution lands.

A note on the generated README

README.md is generated, not hand-written. It is built from the survey's references.bib plus a curated supplement, so an edit made directly to README.md will be overwritten the next time it is regenerated.

That does not mean your PR is unwelcome. Open it against README.md as normal: the maintainer folds accepted entries into the source that generates the file, which is what makes the change stick.

Sibling list

For the field-wide map of LLM agents rather than the loop itself, see Awesome LLM Agent Papers. A paper about agent memory, applications, or multi-agent systems in general probably belongs there instead. Papers that are canonical for both lists appear in both, deliberately.