https://arxiv.org/abs/2411.12580 Skip to main content Cornell University We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate arxiv logo > cs > arXiv:2411.12580 [ ] Help | Advanced Search [All fields ] Search arXiv logo Cornell University Logo [ ] GO quick links * Login * Help Pages * About Computer Science > Computation and Language arXiv:2411.12580 (cs) [Submitted on 19 Nov 2024] Title:Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models Authors:Laura Ruis, Maximilian Mozes, Juhan Bae, Siddhartha Rao Kamalakara, Dwarak Talupuru, Acyr Locatelli, Robert Kirk, Tim Rocktaschel, Edward Grefenstette, Max Bartolo View a PDF of the paper titled Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models, by Laura Ruis and 9 other authors View PDF HTML (experimental) Abstract:The capabilities and limitations of Large Language Models have been sketched out in great detail in recent years, providing an intriguing yet conflicting picture. On the one hand, LLMs demonstrate a general ability to solve problems. On the other hand, they show surprising reasoning gaps when compared to humans, casting doubt on the robustness of their generalisation strategies. The sheer volume of data used in the design of LLMs has precluded us from applying the method traditionally used to measure generalisation: train-test set separation. To overcome this, we study what kind of generalisation strategies LLMs employ when performing reasoning tasks by investigating the pretraining data they rely on. For two models of different sizes (7B and 35B) and 2.5B of their pretraining tokens, we identify what documents influence the model outputs for three simple mathematical reasoning tasks and contrast this to the data that are influential for answering factual questions. We find that, while the models rely on mostly distinct sets of data for each factual question, a document often has a similar influence across different reasoning questions within the same task, indicating the presence of procedural knowledge. We further find that the answers to factual questions often show up in the most influential data. However, for reasoning questions the answers usually do not show up as highly influential, nor do the answers to the intermediate reasoning steps. When we characterise the top ranked documents for the reasoning questions qualitatively, we confirm that the influential documents often contain procedural knowledge, like demonstrating how to obtain a solution using formulae or code. Our findings indicate that the approach to reasoning the models use is unlike retrieval, and more like a generalisable strategy that synthesises procedural knowledge from documents doing a similar form of reasoning. Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG) Cite as: arXiv:2411.12580 [cs.CL] (or arXiv:2411.12580v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2411.12580 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Laura Ruis [view email] [v1] Tue, 19 Nov 2024 15:47:12 UTC (22,020 KB) Full-text links: Access Paper: View a PDF of the paper titled Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models, by Laura Ruis and 9 other authors * View PDF * HTML (experimental) * TeX Source * Other Formats view license Current browse context: cs.CL < prev | next > new | recent | 2024-11 Change to browse by: cs cs.LG References & Citations * NASA ADS * Google Scholar * Semantic Scholar a export BibTeX citation Loading... BibTeX formatted citation x [loading... ] Data provided by: Bookmark BibSonomy logo Reddit logo (*) Bibliographic Tools Bibliographic and Citation Tools [ ] Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) [ ] Connected Papers Toggle Connected Papers (What is Connected Papers?) [ ] Litmaps Toggle Litmaps (What is Litmaps?) [ ] scite.ai Toggle scite Smart Citations (What are Smart Citations?) ( ) Code, Data, Media Code, Data and Media Associated with this Article [ ] alphaXiv Toggle alphaXiv (What is alphaXiv?) [ ] Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) [ ] DagsHub Toggle DagsHub (What is DagsHub?) [ ] GotitPub Toggle Gotit.pub (What is GotitPub?) [ ] Huggingface Toggle Hugging Face (What is Huggingface?) [ ] Links to Code Toggle Papers with Code (What is Papers with Code?) [ ] ScienceCast Toggle ScienceCast (What is ScienceCast?) ( ) Demos Demos [ ] Replicate Toggle Replicate (What is Replicate?) [ ] Spaces Toggle Hugging Face Spaces (What is Spaces?) [ ] Spaces Toggle TXYZ.AI (What is TXYZ.AI?) ( ) Related Papers Recommenders and Search Tools [ ] Link to Influence Flower Influence Flower (What are Influence Flowers?) [ ] Core recommender toggle CORE Recommender (What is CORE?) * Author * Venue * Institution * Topic ( ) About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) * About * Help * Click here to contact arXiv Contact * Click here to subscribe Subscribe * Copyright * Privacy Policy * Web Accessibility Assistance * arXiv Operational Status Get status notifications via email or slack