# Natural Language Inference on HANS (val) Source: [Natural Language Inference on HANS (val)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:28:31.999Z Evaluation of Natural Language Inference performance on the validation split of the HANS (Heuristic Analysis for NLI Systems) dataset, measured by Accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Natural Language Inference | | Dataset (including any stated split) | HANS (val) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 28 of 28 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | ConfReg | 84.3 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-PGI | 84.17 | Not recorded | 68fa6da81e209131cea3b318 | | ERM | 84.12 | Not recorded | 68fa6da81e209131cea3b318 | | ERM | 84.12 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-REx | 84.1 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-IRM | 84.01 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-IRM | 84.01 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-REx | 84 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-PGI | 83.94 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-REx | 83.91 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-IRM | 83.82 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-PGI | 83.78 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-cMMD | 83.56 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-cMMD | 83.38 | Not recorded | 68fa6da81e209131cea3b318 | | EIIL-cMMD | 83.22 | Not recorded | 68fa6da81e209131cea3b318 | | PoE | 82.8 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-PGI | 82.79 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-PGI | 82.79 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-IRM | 82.75 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-cMMD | 82.74 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-REx | 82.68 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-IRM | 82.67 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-cMMD | 82.61 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-REx | 82.59 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-IRM | 82.56 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-REx | 82.56 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-cMMD | 82.39 | Not recorded | 68fa6da81e209131cea3b318 | | SCILL-PGI | 81.69 | Not recorded | 68fa6da81e209131cea3b318 | ### ConfReg Accuracy: 84.3. Result ID: 698fb50b3a6c314849fb4806. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"-"}. This record does not assert independent reproduction. ### EIIL-PGI Accuracy: 84.17. Result ID: 698fb50b3a6c314849fb4810. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"PGI"}. This record does not assert independent reproduction. ### ERM Accuracy: 84.12. Result ID: 698fb50b3a6c314849fb4803. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"-"}. This record does not assert independent reproduction. ### ERM Accuracy: 84.12. Result ID: 698fb50b3a6c314849fb4804. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"-"}. This record does not assert independent reproduction. ### EIIL-REx Accuracy: 84.1. Result ID: 698fb50b3a6c314849fb480a. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"REx"}. This record does not assert independent reproduction. ### EIIL-IRM Accuracy: 84.01. Result ID: 698fb50b3a6c314849fb4807. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"IRM"}. This record does not assert independent reproduction. ### EIIL-IRM Accuracy: 84.01. Result ID: 698fb50b3a6c314849fb4809. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"IRM"}. This record does not assert independent reproduction. ### EIIL-REx Accuracy: 84. Result ID: 698fb50b3a6c314849fb480c. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"REx"}. This record does not assert independent reproduction. ### EIIL-PGI Accuracy: 83.94. Result ID: 698fb50b3a6c314849fb4812. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"PGI"}. This record does not assert independent reproduction. ### EIIL-REx Accuracy: 83.91. Result ID: 698fb50b3a6c314849fb480b. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"REx"}. This record does not assert independent reproduction. ### EIIL-IRM Accuracy: 83.82. Result ID: 698fb50b3a6c314849fb4808. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"IRM"}. This record does not assert independent reproduction. ### EIIL-PGI Accuracy: 83.78. Result ID: 698fb50b3a6c314849fb4811. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"PGI"}. This record does not assert independent reproduction. ### EIIL-cMMD Accuracy: 83.56. Result ID: 698fb50b3a6c314849fb480d. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"cMMD"}. This record does not assert independent reproduction. ### EIIL-cMMD Accuracy: 83.38. Result ID: 698fb50b3a6c314849fb480f. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"cMMD"}. This record does not assert independent reproduction. ### EIIL-cMMD Accuracy: 83.22. Result ID: 698fb50b3a6c314849fb480e. [Reporting paper]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"cMMD"}. This record does not assert independent reproduction. ### PoE Accuracy: 82.8. Result ID: 698fb50b3a6c314849fb4805. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"-"}. This record does not assert independent reproduction. ### SCILL-PGI Accuracy: 82.79. Result ID: 698fb50b3a6c314849fb481c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"PGI"}. This record does not assert independent reproduction. ### SCILL-PGI Accuracy: 82.79. Result ID: 698fb50b3a6c314849fb481e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"PGI"}. This record does not assert independent reproduction. ### SCILL-IRM Accuracy: 82.75. Result ID: 698fb50b3a6c314849fb4813. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"IRM"}. This record does not assert independent reproduction. ### SCILL-cMMD Accuracy: 82.74. Result ID: 698fb50b3a6c314849fb4819. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"cMMD"}. This record does not assert independent reproduction. ### SCILL-REx Accuracy: 82.68. Result ID: 698fb50b3a6c314849fb4816. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"ID","Penalty":"REx"}. This record does not assert independent reproduction. ### SCILL-IRM Accuracy: 82.67. Result ID: 698fb50b3a6c314849fb4815. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"IRM"}. This record does not assert independent reproduction. ### SCILL-cMMD Accuracy: 82.61. Result ID: 698fb50b3a6c314849fb481b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"cMMD"}. This record does not assert independent reproduction. ### SCILL-REx Accuracy: 82.59. Result ID: 698fb50b3a6c314849fb4817. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"REx"}. This record does not assert independent reproduction. ### SCILL-IRM Accuracy: 82.56. Result ID: 698fb50b3a6c314849fb4814. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"IRM"}. This record does not assert independent reproduction. ### SCILL-REx Accuracy: 82.56. Result ID: 698fb50b3a6c314849fb4818. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"TEV","Penalty":"REx"}. This record does not assert independent reproduction. ### SCILL-cMMD Accuracy: 82.39. Result ID: 698fb50b3a6c314849fb481a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"cMMD"}. This record does not assert independent reproduction. ### SCILL-PGI Accuracy: 81.69. Result ID: 698fb50b3a6c314849fb481d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Selection Strategy":"Oracle","Penalty":"PGI"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.