{"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/texttt-metabench-a-sparse-benchmark-to","title":"$\\texttt{metabench}$ -- A Sparse Benchmark to Measure General Ability in Large Language Models","arxiv_id":"2407.12844","date":"2024-07-04","proceeding":null,"authors":["Alex Kipnis","Konstantinos Voudouris","Luca M. Schulze Buschoff","Eric Schulz"],"abstract":"Large Language Models (LLMs) vary in their abilities on a range of tasks. Initiatives such as the $\\texttt{Open LLM Leaderboard}$ aim to quantify these differences with several large benchmarks (sets of test items to which an LLM can respond either correctly or incorrectly). However, high correlations within and between benchmark scores suggest that (1) there exists a small set of common underlying abilities that these benchmarks measure, and (2) items tap into redundant information and the benchmarks may thus be considerably compressed. We use data from $n > 5000$ LLMs to identify the most informative items of six benchmarks, ARC, GSM8K, HellaSwag, MMLU, TruthfulQA and WinoGrande (with $d=28,632$ items in total). From them we distill a sparse benchmark, $\\texttt{metabench}$, that has less than $3\\%$ of the original size of all six benchmarks combined. This new sparse benchmark goes beyond point scores by yielding estimators of the underlying benchmark-specific abilities. We show that these estimators (1) can be used to reconstruct each original $\\textit{individual}$ benchmark score with, on average, $1.5\\%$ root mean square error (RMSE), (2) reconstruct the original $\\textit{total}$ score with $0.8\\%$ RMSE, and (3) have a single underlying common factor whose Spearman correlation with the total score is $r = 0.93$.","url_abs":"https://arxiv.org/abs/2407.12844v1","url_pdf":"https://arxiv.org/pdf/2407.12844v1.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":"texttt-metabench-a-sparse-benchmark-to","repo_url":"https://github.com/adkipnis/metabench","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"arc","task_name":"ARC"},{"task_slug":"gsm8k","task_name":"GSM8K"},{"task_slug":"hellaswag","task_name":"HellaSwag"},{"task_slug":"mmlu","task_name":"MMLU"},{"task_slug":"truthfulqa","task_name":"TruthfulQA"},{"task_slug":"winogrande","task_name":"Winogrande"}],"methods":[{"method_slug":"set","method_name":"SET"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2407.12844","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2407.12844"}},"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":"deterministic:regex_extraction","url":"https://github.com/adkipnis/metabench","reach":{"status":"ok","spdx":"NOASSERTION"}}],"summary":{"ran":8,"unverified":1},"by_repo_kind":{"official":{"samples":9,"ran":8,"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":9,"samples":[{"code_sha256_prefix":"da9b84213864fc42","entry":"collect_shortprompts","repo":"adkipnis/metabench","repo_kind":"official","path":"items/create_metabench.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/items/create_metabench.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"da9b84213864fc42"}},{"code_sha256_prefix":"b67da21279f6e00f","entry":"collect_shortprompts_wg","repo":"adkipnis/metabench","repo_kind":"official","path":"items/create_metabench.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/items/create_metabench.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"b67da21279f6e00f"}},{"code_sha256_prefix":"781799876ce99d3e","entry":"data_checker","repo":"adkipnis/metabench","repo_kind":"official","path":"items/create_metabench.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/items/create_metabench.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"781799876ce99d3e"}},{"code_sha256_prefix":"2dd93c6df901fe64","entry":"infoplot","repo":"adkipnis/metabench","repo_kind":"official","path":"figures/overview/paperplots.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/figures/overview/paperplots.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"2dd93c6df901fe64"}},{"code_sha256_prefix":"e00a2b91fe0657b6","entry":"parameterplot","repo":"adkipnis/metabench","repo_kind":"official","path":"figures/overview/paperplots.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/figures/overview/paperplots.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"e00a2b91fe0657b6"}},{"code_sha256_prefix":"15fb9133086b4ae9","entry":"process_gsm8k","repo":"adkipnis/metabench","repo_kind":"official","path":"finetuning/mb-finetune.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/finetuning/mb-finetune.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"15fb9133086b4ae9"}},{"code_sha256_prefix":"da871d87f24ae38b","entry":"process_hellaswag","repo":"adkipnis/metabench","repo_kind":"official","path":"finetuning/mb-finetune.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/finetuning/mb-finetune.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"da871d87f24ae38b"}},{"code_sha256_prefix":"231316e10eecda23","entry":"sbatchGen","repo":"adkipnis/metabench","repo_kind":"official","path":"scraping/sbatch_generator.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/scraping/sbatch_generator.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"231316e10eecda23"}},{"code_sha256_prefix":"06e13c4f1a02aca1","entry":"process_arc","repo":"adkipnis/metabench","repo_kind":"official","path":"finetuning/mb-finetune.py","file_url":"https://github.com/adkipnis/metabench/blob/HEAD/finetuning/mb-finetune.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"06e13c4f1a02aca1"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}