{"about":{"site":"https://codewithpapers.app","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.","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"},"url":"/paper/the-surprising-effectiveness-of-test-time","title":"The Surprising Effectiveness of Test-Time Training for Few-Shot Learning","arxiv_id":"2411.07279","date":"2024-11-11","proceeding":null,"authors":["Ekin Akyürek","Mehul Damani","Adam Zweiger","Linlu Qiu","Han Guo","Jyothish Pari","Yoon Kim","Jacob Andreas"],"abstract":"Language models (LMs) have shown impressive performance on tasks within their training distribution, but often struggle with structurally novel tasks even when given a small number of in-context task examples. We investigate the effectiveness of test-time training (TTT) -- temporarily updating model parameters during inference using a loss derived from input data -- as a mechanism for improving LMs' reasoning and few-shot learning capabilities. On the Abstraction and Reasoning Corpus (ARC), performing TTT with in-context examples yields up to $6\\times$ higher accuracy compared to fine-tuned baselines -- reaching $53.0\\%$ on the public validation set with an 8B-parameter LM and $61.9\\%$ when ensembled with program-synthesis methods, matching average human performance. On BIG-Bench Hard (BBH), TTT on in-context examples surpasses standard few-shot prompting in the $10$-shot setting by $7.3$ percentage points ($50.5\\%$ to $57.8\\%$). Our findings highlight the limitations of in-context learning for novel tasks and demonstrate the potential of test-time training to enhance language model adaptability.","url_abs":"https://arxiv.org/abs/2411.07279v2","url_pdf":"https://arxiv.org/pdf/2411.07279v2.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":"the-surprising-effectiveness-of-test-time","repo_url":"https://github.com/ekinakyurek/marc","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"arc","task_name":"ARC"},{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"in-context-learning","task_name":"In-Context Learning"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"program-synthesis","task_name":"Program Synthesis"}],"methods":[{"method_slug":"base","method_name":"BASE"},{"method_slug":"set","method_name":"SET"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2411.07279","atlas_url":"https://app.syntology.ai/?focus=2411.07279","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2411.07279"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ekinakyurek/marc","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":10},"by_repo_kind":{"official":{"samples":10,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"b90ccea3c67381c1","entry":"display_messages","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/messagers.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/messagers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b90ccea3c67381c1"}},{"code_sha256_prefix":"540fdd9083bbe299","entry":"get_all_type_of_votingsv2","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/voting.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/voting.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"540fdd9083bbe299"}},{"code_sha256_prefix":"14aaa6991d622219","entry":"np_lru_cache","repo":"ekinakyurek/marc","repo_kind":"official","path":"utils/np_cache.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/utils/np_cache.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"14aaa6991d622219"}},{"code_sha256_prefix":"96555a8ebf7f060b","entry":"parse_numpy_from_str","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/representers.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/representers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"96555a8ebf7f060b"}},{"code_sha256_prefix":"43bb60cb4cdce0cb","entry":"read_tasks_from_folder","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/arc.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/arc.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"43bb60cb4cdce0cb"}},{"code_sha256_prefix":"9700fc1ad6d2c336","entry":"row_base_majority_voting","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/voting.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/voting.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9700fc1ad6d2c336"}},{"code_sha256_prefix":"942577fea41d2316","entry":"setup_lm_api","repo":"ekinakyurek/marc","repo_kind":"official","path":"utils/lm_api.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/utils/lm_api.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"942577fea41d2316"}},{"code_sha256_prefix":"95bd0d71a136300c","entry":"to_list","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/arc.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/arc.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"95bd0d71a136300c"}},{"code_sha256_prefix":"77f33bc5b258bb01","entry":"to_tuple","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/arc.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/arc.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"77f33bc5b258bb01"}},{"code_sha256_prefix":"9bf200de6989aff4","entry":"vote","repo":"ekinakyurek/marc","repo_kind":"official","path":"arclib/voting.py","file_url":"https://github.com/ekinakyurek/marc/blob/HEAD/arclib/voting.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9bf200de6989aff4"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}