Submission Instructions

Please see the Call For Papers for full instructions. We highlight a few key reminders here.

Submit an abstract by March 26 AOE (11:59pm UTC-12)

Please submit an abstract by the abstract deadline. Uploading a PDF at this time is optional.

All authors must have complete OpenReview profiles with up-to-date affiliation information to enable conflict checks.

After this deadline, you may make minor changes to the title/abstract, but not major changes. Making major changes is grounds for desk rejection. We begin the process of AC and reviewer assignment once abstracts are in, so a substantial change in the content (e.g., submitting a completely different paper) would result in the paper not being reviewed appropriately.

Follow the template

Please note that COLM uses a distinct template from other venues. It's available at: https://github.com/COLM-org/Template/releases/tag/2026

Be careful if you're resubmitting a paper! Make sure it continues to conform to the template, and ensure that any imported packages (especially geometry) do not alter the template parameters.

Disclose use of LLMs according to the policy

The CFP describes that:

In research, use of LLMs should be disclosed in the submitted paper, except for minor usage in paper writing or code implementation.
- Must be disclosed: using an LLM to originate research ideas, using an LLM to write original content in the paper (including references), using an LLM to generate data or plots, using an LLM for evaluation
- Okay not to disclose: using automated grammar-checking tools, feeding a paragraph to an LLM and asking it to fix typos, programming with a tool like CoPilot or Cursor.

Make sure this disclosure is clear in your work.

In addition, usage of LLMs in a way that undermines scientific credibility of the paper will lead to rejection. For instance, clear use of an LLM that results in hallucinated content in one part of the paper casts doubt on the claims of the entire paper.

Finally, we expect our reviewers to disclose any usage of LLMs and to fully stand behind the content of their reviews.