# Scientific Natural Language Inference on SciNLI 1.0 (test) Source: [Scientific Natural Language Inference on SciNLI 1.0 (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:51:26.303Z Evaluation of Scientific Natural Language Inference performance on the test split of the SciNLI 1.0 dataset, measured by Macro F1 and 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 | Scientific Natural Language Inference | | Dataset (including any stated split) | SciNLI 1.0 (test) | | Metric | Macro F1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 32 of 32 | | 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 - [Macro F1](): Higher is better; unit: PERCENT. - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Macro F1 | Result date | Report paper | | --- | --- | --- | --- | | WCT-CV | 79.62 | Not recorded | 68fa73c21e209131cea5e4ed | | WCT-CC | 79.12 | Not recorded | 68fa73c21e209131cea5e4ed | | CO-TEACHING | 78.72 | Not recorded | 68fa73c21e209131cea5e4ed | | WCT-CV | 78.55 | Not recorded | 68fa73c21e209131cea5e4ed | | WCT-CC | 78.35 | Not recorded | 68fa73c21e209131cea5e4ed | | DS-LABELED (101K) | 78.08 | Not recorded | 68fa73c21e209131cea5e4ed | | CO-TEACHING | 78.06 | Not recorded | 68fa73c21e209131cea5e4ed | | DS-LABELED (101K) | 77.53 | Not recorded | 68fa73c21e209131cea5e4ed | | CO-TEACHING+ | 76.47 | Not recorded | 68fa73c21e209131cea5e4ed | | CO-TEACHING+ | 76.27 | Not recorded | 68fa73c21e209131cea5e4ed | | DBST | 73.74 | Not recorded | 68fa73c21e209131cea5e4ed | | DBST | 73.61 | Not recorded | 68fa73c21e209131cea5e4ed | | SOFTMATCH | 71.8 | Not recorded | 68fa73c21e209131cea5e4ed | | FLEXMATCH | 71.64 | Not recorded | 68fa73c21e209131cea5e4ed | | FIXMATCH | 71.27 | Not recorded | 68fa73c21e209131cea5e4ed | | CO-TRAINING | 70.35 | Not recorded | 68fa73c21e209131cea5e4ed | | BT (H-LABELED 2K) | 70.23 | Not recorded | 68fa73c21e209131cea5e4ed | | FS (H-LABELED 2K) | 70.13 | Not recorded | 68fa73c21e209131cea5e4ed | | SOFTMATCH | 68.77 | Not recorded | 68fa73c21e209131cea5e4ed | | FLEXMATCH | 68.6 | Not recorded | 68fa73c21e209131cea5e4ed | | CO-TRAINING | 68.29 | Not recorded | 68fa73c21e209131cea5e4ed | | FIXMATCH | 68.21 | Not recorded | 68fa73c21e209131cea5e4ed | | BT (H-LABELED 2K) | 68.12 | Not recorded | 68fa73c21e209131cea5e4ed | | BT (H-LABELED 1K) | 67.73 | Not recorded | 68fa73c21e209131cea5e4ed | | FS (H-LABELED 1K) | 67.7 | Not recorded | 68fa73c21e209131cea5e4ed | | FS (H-LABELED 2K) | 67.46 | Not recorded | 68fa73c21e209131cea5e4ed | | DS-LABELED (2K) | 67.01 | Not recorded | 68fa73c21e209131cea5e4ed | | FS (H-LABELED 1K) | 65.31 | Not recorded | 68fa73c21e209131cea5e4ed | | DS-LABELED (2K) | 64.74 | Not recorded | 68fa73c21e209131cea5e4ed | | BT (H-LABELED 1K) | 64.13 | Not recorded | 68fa73c21e209131cea5e4ed | | DS-LABELED (1K) | 63.29 | Not recorded | 68fa73c21e209131cea5e4ed | | DS-LABELED (1K) | 61.21 | Not recorded | 68fa73c21e209131cea5e4ed | ### WCT-CV Macro F1: 79.62. Result ID: 6993218e3a6c31484913ed61. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### WCT-CC Macro F1: 79.12. Result ID: 6993218e3a6c31484913ed5f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### CO-TEACHING Macro F1: 78.72. Result ID: 6993218e3a6c31484913ed5b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### WCT-CV Macro F1: 78.55. Result ID: 6993218e3a6c31484913ed60. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### WCT-CC Macro F1: 78.35. Result ID: 6993218e3a6c31484913ed5e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### DS-LABELED (101K) Macro F1: 78.08. Result ID: 6993218e3a6c31484913ed4f. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### CO-TEACHING Macro F1: 78.06. Result ID: 6993218e3a6c31484913ed5a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### DS-LABELED (101K) Macro F1: 77.53. Result ID: 6993218e3a6c31484913ed4e. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### CO-TEACHING+ Macro F1: 76.47. Result ID: 6993218e3a6c31484913ed5d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### CO-TEACHING+ Macro F1: 76.27. Result ID: 6993218e3a6c31484913ed5c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### DBST Macro F1: 73.74. Result ID: 6993218e3a6c31484913ed51. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### DBST Macro F1: 73.61. Result ID: 6993218e3a6c31484913ed50. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### SOFTMATCH Macro F1: 71.8. Result ID: 6993218e3a6c31484913ed57. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### FLEXMATCH Macro F1: 71.64. Result ID: 6993218e3a6c31484913ed55. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### FIXMATCH Macro F1: 71.27. Result ID: 6993218e3a6c31484913ed53. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### CO-TRAINING Macro F1: 70.35. Result ID: 6993218e3a6c31484913ed59. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### BT (H-LABELED 2K) Macro F1: 70.23. Result ID: 6993218e3a6c31484913ed4d. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### FS (H-LABELED 2K) Macro F1: 70.13. Result ID: 6993218e3a6c31484913ed4b. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### SOFTMATCH Macro F1: 68.77. Result ID: 6993218e3a6c31484913ed56. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### FLEXMATCH Macro F1: 68.6. Result ID: 6993218e3a6c31484913ed54. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### CO-TRAINING Macro F1: 68.29. Result ID: 6993218e3a6c31484913ed58. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### FIXMATCH Macro F1: 68.21. Result ID: 6993218e3a6c31484913ed52. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### BT (H-LABELED 2K) Macro F1: 68.12. Result ID: 6993218e3a6c31484913ed4c. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### BT (H-LABELED 1K) Macro F1: 67.73. Result ID: 6993218e3a6c31484913ed47. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### FS (H-LABELED 1K) Macro F1: 67.7. Result ID: 6993218e3a6c31484913ed45. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### FS (H-LABELED 2K) Macro F1: 67.46. Result ID: 6993218e3a6c31484913ed4a. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### DS-LABELED (2K) Macro F1: 67.01. Result ID: 6993218e3a6c31484913ed49. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### FS (H-LABELED 1K) Macro F1: 65.31. Result ID: 6993218e3a6c31484913ed44. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### DS-LABELED (2K) Macro F1: 64.74. Result ID: 6993218e3a6c31484913ed48. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### BT (H-LABELED 1K) Macro F1: 64.13. Result ID: 6993218e3a6c31484913ed46. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. ### DS-LABELED (1K) Macro F1: 63.29. Result ID: 6993218e3a6c31484913ed43. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### DS-LABELED (1K) Macro F1: 61.21. Result ID: 6993218e3a6c31484913ed42. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"SciBERT"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.