Efficient Benchmarking (of Language Models)
Open accessAbstract
Yotam Perlitz, Elron Bandel, Ariel Gera, Ofir Arviv, Liat Ein-Dor, Eyal Shnarch, Noam Slonim, Michal Shmueli-Scheuer, Leshem Choshen. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Common heuristics such as selecting 1K or fewer examples tend to produce highly uncertain estimates, often exceeding ±5.5 points.
“Second, common heuristics (Perlitz et al., 2024) - such as selecting 1K or fewer examples - tend to produce highly uncertain estimates, often exceeding ± 5 . 5 points.”
From Stop Guessing When to Stop Testing: Efficient Model Evaluation with Just Enough Data (2026) · cites this paper as Perlitz et al., 2024
Section 5.3 (Examples; page not available) reports high reliability with a small number of examples and says model ranks remain quite stable as the number of examples changes. That runs contrary to the claim that selecting 1K or fewer examples tends to yield highly uncertain estimates. The passage does not report uncertainty exceeding ±5.5 points. (LLM, conf=0.84)
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredPractitioners often reduce costs by using fewer samples, but these heuristic approaches lack statistical guarantees.
“While practitioners often reduce costs by using fewer samples (Perlitz et al., 2024; Polo et al., 2024a; Fogliato et al., 2024; Zhao et al., 2024), these heuristic approaches lack statistical guarantees.”
From Stop Guessing When to Stop Testing: Efficient Model Evaluation with Just Enough Data (2026) · cites this paper as Perlitz et al., 2024
Full benchmarks should be larger and include large datasets to allow for reliable results.
“A secondary finding is that full benchmarks should be larger (echoing (Perlitz et al., 2024)) and include large datasets to allow for reliable results.”
From Stop Guessing When to Stop Testing: Efficient Model Evaluation with Just Enough Data (2026) · cites this paper as Perlitz et al., 2024
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
The publisher hasn’t deposited this paper’s reference list.
Research works printed but not linked: 4
- BIG bench authors. 2023. Beyond the imitation game: Quantifying and extrapolating the capabilities of lan- guage models. Transactions on Machine Learning Research.
- Leshem Choshen and Omri Abend. 2018a. Automatic metric validation for grammatical error correction. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 1372-1382, Melbourne, Aus- tralia. Association for Computational Linguistics. · DOI
- Rebecca Knowles. 2021. On the stability of system rankings at WMT. In Proceedings of the Sixth Con- ference on Machine Translation, pages 464-477, On- line. Association for Computational Linguistics.
- Maxime Peyrard, Wei Zhao, Steffen Eger, and Robert West. 2021. Better than average: Paired evaluation of NLP systems. In Proceedings of the 59th Annual Meeting of the Association for Computational Lin- guistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages 2301-2315, Online. Association for Computational Linguistics. · DOI
Cited by
Sorted by citations
State of What Art? A Call for Multi-Prompt LLM Evaluation
Mizrahi, M., Kaplan, G., Malkin, D., Dror, R., Shahaf, D., & Stanovsky, G. (2024) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00681 · Cited by 113 · see citing works
Automated Generation of Challenging Multiple-Choice Questions for Vision Language Model Evaluation
Zhang, Y., Su, Y., Liu, Y., Wang, X., Burgess, J., Sui, E., Wang, C., Aklilu, J. G., Lozano, A., Wei, A., et al. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.02754 · Cited by 6 · see citing works
Holmes ⌕ A Benchmark to Assess the Linguistic Competence of Language Models
Waldis, A., Waldis, A., Perlitz, Y., Choshen, L., Choshen, L., Hou, Y., & Gurevych, I. (2024) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00718 · Cited by 9 · see citing works
Self-citationUser-defined trade-offs in LLM benchmarking: balancing accuracy, scale, and sustainability
Gjorgjevikj, A., Nikolikj, A., Seljak, B. K., & Eftimov, T. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.114405 · Cited by 5 · see citing works
Multi-Criteria Evaluation of Large Language Models (LLMs): Balancing Performance and Security
Colares, D. M., Pinheiro, P., & Holanda, R. (2026) · IEEE Access · doi:10.1109/access.2026.3665546 · Cited by 1 · see citing works
Liu, F., Liu, Y., Sun, R., Luo, L., & Guo, S. (2026) · arXiv · doi:10.48550/arxiv.2609.17394
Stop Guessing When to Stop Testing: Efficient Model Evaluation with Just Enough Data
Arviv, O., Greenewald, K., Perlitz, Y., Mulian, H., Shmueli-Scheuer, M., & Choshen, L. (2026) · Findings of the Association for Computational Linguistics: ACL 2026 · doi:10.18653/v1/2026.findings-acl.43
Self-citationEfficient Benchmarking via Bias-Bounded Subset Selection
Zhuang, Y., Yu, J., Liu, Q., Sun, Y., Li, J., Huang, Z., & Chen, E. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3598031
Concept Control for LLM Safety Using Radial Basis Function Representations
Amos, M., Song, Y., & Pagnucco, M. (2025) · Lecture notes in computer science · doi:10.1007/978-981-95-4969-6_17
- Authors
- Perlitz, Y., Bandel, E., Gera, A., Arviv, O., Ein-Dor, L., Shnarch, E., Slonim, N., Shmueli-Scheuer, M., & Choshen, L.
- Venue
- Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)
- Published
- 2024
- Pages
- 2519-2536
- OpenAlex
- W4401042935
- Cited by
- 7
- References
- 0
- Citations by year
- 2026: 1 · 2025: 4 · 2024: 2
- FWCI
- 2.02
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Information Systems › Information Retrieval and Search Behavior
- Keywords
- language model benchmarking, language model evaluation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
