# LitSumm LitSumm uses large language models to write short summaries of what is known about non-coding RNAs from the scientific literature. Each summary is built from the open access articles that [LitScan](litscan.md) links to the RNA, and cites the articles it draws on. The approach is described in [LitSumm: large language models for literature summarization of noncoding RNAs](https://doi.org/10.1093/database/baaf006). [Browse all sequences with LitSumm summaries →](https://rnacentral.org/search?q=has_litsumm:%22True%22) ## Summaries on RNAcentral pages Summaries appear near the top of a sequence page, below the RNA type, under an **AI-generated content** notice. A sequence can have more than one summary, one for each name it is known by in the literature. Each article cited in a summary is linked to Europe PMC. For example, see the summary for [human MIR944](https://rnacentral.org/rna/URS0000663F9C/9606). [Gene pages](genes.md) show the summaries of the transcripts in the gene. ## How summaries are made For each RNA, a language model reads the relevant passages of the articles found by LitScan and writes a summary that cites each statement. The summary is then checked automatically against the articles and revised where needed. ## Limitations Summaries are generated automatically and may not be completely accurate or comprehensive. Known issues include: - missing or incorrectly formatted references; - similar statements from different articles merged incorrectly; - inaccurate statements. If you notice a problem with a summary, use the **Give feedback** link next to it.