{"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/arxiv-2607-04429","title":"evalci: A Python Library for Statistically Rigorous Comparison of Language Model Evaluations","arxiv_id":"2607.04429","date":"2026-07-05","proceeding":null,"authors":["Shreyas K Chandrahas"],"abstract":"The dominant practice in language model evaluation is to report a single accuracy number per model and declare the higher one better, without testing whether the gap could plausibly be sampling noise. On benchmarks of a few thousand items, and under temperature sampling where a model can differ from itself run to run by more than the reported gap between models, this practice routinely overstates confidence in headline claims. The statistical machinery to fix this -- confidence intervals, paired significance tests, power analysis, clustered standard errors, multiple-comparison correction -- is well established, but no standard, pip-installable tool packages it in the shape an evaluation actually takes: a per-item results table. We present evalci, a pure-Python library (numpy/scipy/pandas only) that turns a per-item results table into a publication-ready claim -- e.g., \"Model A beats Model B, $Δ=3.1$ pts, 95% CI [1.2, 5.0], paired permutation $p=0.002$, $n=1{,}319$\" -- in one function call, with adapters for lm-evaluation-harness and HELM output. Every routine is validated against an independent reference (statsmodels, or brute-force exact enumeration) rather than only against itself. As a case study, we re-analyze a public comparison of nine language models' MMLU accuracy and find that 3 of the 8 adjacent leaderboard-rank gaps are not statistically significant after correcting for the 36 pairwise comparisons the ranking implies. evalci is available at https://pypi.org/project/evalci/ (source: https://github.com/Shreyaskc/evalci, DOI: https://doi.org/10.5281/zenodo.21201815)","url_abs":"https://arxiv.org/abs/2607.04429","url_pdf":"https://arxiv.org/pdf/2607.04429","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2607.04429","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2607.04429"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/Shreyaskc/evalci","reach":null}],"summary":{"ran":2,"ran_draft_wrong":2,"ran_honours":3,"ran_fixture":6,"unverified":1},"by_repo_kind":{"found_in_text":{"samples":14,"ran":13,"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":"6504ccb986870a77","entry":"CIResult","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6504ccb986870a77"}},{"code_sha256_prefix":"891cb21c235fa770","entry":"CompareResult","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"891cb21c235fa770"}},{"code_sha256_prefix":"da9874b0603c50e9","entry":"bootstrap_interval","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"da9874b0603c50e9"}},{"code_sha256_prefix":"6219547495fa7208","entry":"ci","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6219547495fa7208"}},{"code_sha256_prefix":"f03ac473a85bde77","entry":"clopper_pearson_interval","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f03ac473a85bde77"}},{"code_sha256_prefix":"177e24430c39228f","entry":"mcnemar_test","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"177e24430c39228f"}},{"code_sha256_prefix":"620b1f42d4971727","entry":"paired_bootstrap_ci","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"620b1f42d4971727"}},{"code_sha256_prefix":"6e8bffa3e62ee384","entry":"paired_bootstrap_p","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6e8bffa3e62ee384"}},{"code_sha256_prefix":"04fc13d94e614b4e","entry":"paired_permutation_p","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"04fc13d94e614b4e"}},{"code_sha256_prefix":"4b7a577202a14f70","entry":"unpaired_bootstrap_ci","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4b7a577202a14f70"}},{"code_sha256_prefix":"a547f79c157fc13e","entry":"unpaired_bootstrap_p","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a547f79c157fc13e"}},{"code_sha256_prefix":"4555f85b0d6608a2","entry":"unpaired_permutation_p","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4555f85b0d6608a2"}},{"code_sha256_prefix":"53d188c2463155be","entry":"wilson_interval","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"53d188c2463155be"}},{"code_sha256_prefix":"50c0c4ad02a4c7b8","entry":"compare","repo":"Shreyaskc/evalci","repo_kind":"found_in_text","path":"src/evalci/stats.py","file_url":"https://github.com/Shreyaskc/evalci/blob/HEAD/src/evalci/stats.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":"50c0c4ad02a4c7b8"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CL","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}