{"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/training-on-the-test-task-confounds","title":"Training on the Test Task Confounds Evaluation and Emergence","arxiv_id":"2407.07890","date":"2024-07-10","proceeding":null,"authors":["Ricardo Dominguez-Olmedo","Florian E. Dorner","Moritz Hardt"],"abstract":"We study a fundamental problem in the evaluation of large language models that we call training on the test task. Unlike wrongful practices like training on the test data, leakage, or data contamination, training on the test task is not a malpractice. Rather, the term describes a growing set of practices that utilize knowledge about evaluation tasks at training time. We demonstrate that training on the test task confounds both relative model evaluations and claims about emergent capabilities. We argue that the seeming superiority of one model family over another may be explained by a different degree of training on the test task. To this end, we propose an effective method to adjust for the effect of training on the test task on benchmark evaluations. Put simply, to fine-tune each model under comparison on the same task-relevant data prior to evaluation. We then show that instances of emergent behavior disappear gradually as models train on the test task. Our work promotes a new perspective on the evaluation of large language models, with broad implications for benchmarking and the study of emergent capabilities.","url_abs":"https://arxiv.org/abs/2407.07890v3","url_pdf":"https://arxiv.org/pdf/2407.07890v3.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":"training-on-the-test-task-confounds","repo_url":"https://github.com/socialfoundations/training-on-the-test-task","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"benchmarking","task_name":"Benchmarking"},{"task_slug":"language-modelling","task_name":"Language Modelling"}],"methods":[{"method_slug":"set","method_name":"SET"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2407.07890","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2407.07890"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/socialfoundations/training-on-the-test-task","reach":{"status":"ok"}}],"summary":{"ran":3,"unverified":3},"by_repo_kind":{"official":{"samples":6,"ran":3,"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":6,"samples":[{"code_sha256_prefix":"d13b737925c23199","entry":"gsm8k_rows_to_qa","repo":"socialfoundations/training-on-the-test-task","repo_kind":"official","path":"experiments/data_utils.py","file_url":"https://github.com/socialfoundations/training-on-the-test-task/blob/HEAD/experiments/data_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"d13b737925c23199"}},{"code_sha256_prefix":"0730cf1097662eb1","entry":"mmluaux_row_to_qa","repo":"socialfoundations/training-on-the-test-task","repo_kind":"official","path":"experiments/data_utils.py","file_url":"https://github.com/socialfoundations/training-on-the-test-task/blob/HEAD/experiments/data_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"0730cf1097662eb1"}},{"code_sha256_prefix":"f067ba5db73dc40a","entry":"tokenize_dataset","repo":"socialfoundations/training-on-the-test-task","repo_kind":"official","path":"experiments/data_utils.py","file_url":"https://github.com/socialfoundations/training-on-the-test-task/blob/HEAD/experiments/data_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"f067ba5db73dc40a"}},{"code_sha256_prefix":"0eac4ba9658efdb5","entry":"get_job_memory","repo":"socialfoundations/training-on-the-test-task","repo_kind":"official","path":"experiments/model_utils.py","file_url":"https://github.com/socialfoundations/training-on-the-test-task/blob/HEAD/experiments/model_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"0eac4ba9658efdb5"}},{"code_sha256_prefix":"3b85469ad66d264c","entry":"get_model_name","repo":"socialfoundations/training-on-the-test-task","repo_kind":"official","path":"experiments/model_utils.py","file_url":"https://github.com/socialfoundations/training-on-the-test-task/blob/HEAD/experiments/model_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"3b85469ad66d264c"}},{"code_sha256_prefix":"ab5c0b21bd0dceef","entry":"get_n_gpus","repo":"socialfoundations/training-on-the-test-task","repo_kind":"official","path":"experiments/model_utils.py","file_url":"https://github.com/socialfoundations/training-on-the-test-task/blob/HEAD/experiments/model_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ab5c0b21bd0dceef"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}