{"about":{"non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","site":"https://codewithpapers.app","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/paper/a-training-free-length-extrapolation-approach","title":"A Training-Free Length Extrapolation Approach for LLMs: Greedy Attention Logit Interpolation (GALI)","arxiv_id":"2502.02659","date":"2025-02-04","proceeding":null,"authors":["Yan Li","Tianyi Zhang","Zechuan Li","Soyeon Caren Han"],"abstract":"Transformer-based Large Language Models (LLMs) struggle to process inputs exceeding their training context window, with performance degrading due to positional out-of-distribution (O.O.D.) that disrupt attention computations. Existing solutions, fine-tuning and training-free methods, are limited by computational inefficiency, attention logit outliers or loss of local positional information. To address this, we propose Greedy Attention Logit Interpolation (GALI), a training-free length extrapolation method that maximizes the utilization of pretrained positional intervals while avoiding attention logit outliers through attention logit interpolation. The result demonstrates that GALI consistently outperforms state-of-the-art training-free methods. Our findings reveal that LLMs interpret positional intervals unevenly within their training context window, suggesting that extrapolating within a smaller positional interval range yields superior results-even for short-context tasks. GALI represents a significant step toward resolving the positional O.O.D. challenge, enabling more reliable long-text understanding in LLMs. Our implementation of GALI, along with the experiments from our paper, is open-sourced at https://github.com/AcademyCityL/GALI.","url_abs":"https://arxiv.org/abs/2502.02659v1","url_pdf":"https://arxiv.org/pdf/2502.02659v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"a-training-free-length-extrapolation-approach","repo_url":"https://github.com/academycityl/gali","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"long-context-understanding","task_name":"Long-Context Understanding"}],"methods":[{"method_slug":"attention","method_name":"Attention"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/long-context-understanding-on-l-eval","task":"Long-Context Understanding","dataset":"L-Eval","model":"GALI(Llama3-8b-ins-4k-to-16k)","rank_in_archive_order":1,"of":4,"metrics":{"Average Score":"59.21"},"uses_additional_data":false},{"leaderboard":"/sota/long-context-understanding-on-l-eval","task":"Long-Context Understanding","dataset":"L-Eval","model":"GALI(Llama3-8b-ins-4k-to-32k)","rank_in_archive_order":2,"of":4,"metrics":{"Average Score":"59.10"},"uses_additional_data":false},{"leaderboard":"/sota/long-context-understanding-on-l-eval","task":"Long-Context Understanding","dataset":"L-Eval","model":"GALI(Llama3-8b-ins-8k-to-32k)","rank_in_archive_order":3,"of":4,"metrics":{"Average Score":"42.79"},"uses_additional_data":false},{"leaderboard":"/sota/long-context-understanding-on-l-eval","task":"Long-Context Understanding","dataset":"L-Eval","model":"GALI(Llama3-8b-ins-8k-to-16k)","rank_in_archive_order":4,"of":4,"metrics":{"Average Score":"42.32"},"uses_additional_data":false},{"leaderboard":"/sota/long-context-understanding-on-longbench","task":"Long-Context Understanding","dataset":"LongBench","model":"GALI(Llama3-8b-ins-4k-to-16k)","rank_in_archive_order":1,"of":3,"metrics":{"Average Score":"46.22"},"uses_additional_data":false},{"leaderboard":"/sota/long-context-understanding-on-longbench","task":"Long-Context Understanding","dataset":"LongBench","model":"GALI(Llama3-8b-ins-8k-to-32k)","rank_in_archive_order":2,"of":3,"metrics":{"Average Score":"45.38"},"uses_additional_data":false},{"leaderboard":"/sota/long-context-understanding-on-longbench","task":"Long-Context Understanding","dataset":"LongBench","model":"GALI(Llama3-8b-ins-8k-to-16k)","rank_in_archive_order":3,"of":3,"metrics":{"Average Score":"45.17"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers","held":null,"reason":"arxiv_id_not_in_atlas","checked_against":{"atlas_ids":549449,"atlas_ids_sha256":"afd72399dafd9906e9062927f70655952385a3d75488558e1f0484e950ab09b7","read_at":"2026-10-05T01:46:21Z"},"by_repo":[{"repo":"academycityl/gali","kind":"official","samples":0,"ran":0,"instrument":0,"unverified":0}],"by_repo_is":"one row per repository in the archive's code links, Syntology's graph links or the samples; samples 0 means no sample is linked to this paper, a statement about Syntology's coverage, not about the repository: with harvested_for_other_papers present, Syntology harvested the repository for that many other papers, and without it, Syntology harvested nothing from it; placed_by_identical_code counts samples whose link records no repository, placed here because identical code was harvested from this repository for another paper; ran counts executions on a synthesized input and instrument counts failures of Syntology's instrument, not of the code"},"code_links_note":null,"arxiv_metadata":null,"syntology_extracted_results":null,"record_sha256":"c02b4d74f9dd016f9b53cc3fef20b48a9c9d115cb27f0ceba48c4ccf26357cce","record_changed_at":"2026-09-28","record_changed_at_basis":"first_hashed"}