# Natural Language Understanding on GLUE RoBERTa-base (val) Source: [Natural Language Understanding on GLUE RoBERTa-base (val)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:39:27.003Z Evaluation of general language understanding capabilities across the GLUE benchmark suite, specifically reporting results using the RoBERTa-base configuration on the validation split. 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 Understanding | | Dataset (including any stated split) | GLUE RoBERTa-base (val) | | Metric | CoLA Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [CoLA Score](): Higher is better; unit: PERCENT. - [SST-2 Accuracy](): Higher is better; unit: PERCENT. - [MRPC Score](): Higher is better; unit: PERCENT. - [STSB Correlation](): Higher is better; unit: PERCENT. - [QQP Score](): Higher is better; unit: PERCENT. - [MNLI Accuracy](): Higher is better; unit: PERCENT. - [QNLI Accuracy](): Higher is better; unit: PERCENT. - [RTE Accuracy](): Higher is better; unit: PERCENT. - [GLUE Average Score](): Higher is better; unit: PERCENT. - [ADR Score](): Higher is better; unit: SCALAR. - [AMR Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | CoLA Score | Result date | Report paper | | --- | --- | --- | --- | | Individual | 60.18 | Not recorded | 692e7313d6b0dd334293272c | | DTS-T\* | 59.71 | Not recorded | 692e7313d6b0dd334293272c | | DTS-D | 59.66 | Not recorded | 692e7313d6b0dd334293272c | | DTS-T | 59.33 | Not recorded | 692e7313d6b0dd334293272c | | DTS-D\* | 59.26 | Not recorded | 692e7313d6b0dd334293272c | | Twin-Merging | 59.12 | Not recorded | 692e7313d6b0dd334293272c | | SVD | 58.31 | Not recorded | 692e7313d6b0dd334293272c | | FREE-Merging | 54.5 | Not recorded | 692e7313d6b0dd334293272c | | T-Switch | 53.12 | Not recorded | 692e7313d6b0dd334293272c | | TALL-Mask | 45.81 | Not recorded | 692e7313d6b0dd334293272c | | EMR-Merging | 39.96 | Not recorded | 692e7313d6b0dd334293272c | | RegMean | 36.67 | Not recorded | 692e7313d6b0dd334293272c | | Ties-Merging | 20.48 | Not recorded | 692e7313d6b0dd334293272c | | Task-Arithmetic | 18.78 | Not recorded | 692e7313d6b0dd334293272c | | Weight-Averaging | 13.96 | Not recorded | 692e7313d6b0dd334293272c | | DARE | 9.28 | Not recorded | 692e7313d6b0dd334293272c | ### Individual CoLA Score: 60.18. Result ID: 699997175698665ec40b2217. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### DTS-T\* CoLA Score: 59.71. Result ID: 699997175698665ec40b2225. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa","sparsity\_coefficient (r)":"adaptive"}. This record does not assert independent reproduction. ### DTS-D CoLA Score: 59.66. Result ID: 699997175698665ec40b2224. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa","sparsity\_coefficient (r)":0.3}. This record does not assert independent reproduction. ### DTS-T CoLA Score: 59.33. Result ID: 699997175698665ec40b2223. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa","sparsity\_coefficient (r)":0.3}. This record does not assert independent reproduction. ### DTS-D\* CoLA Score: 59.26. Result ID: 699997175698665ec40b2226. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa","sparsity\_coefficient (r)":"adaptive"}. This record does not assert independent reproduction. ### Twin-Merging CoLA Score: 59.12. Result ID: 699997175698665ec40b221e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### SVD CoLA Score: 58.31. Result ID: 699997175698665ec40b221d. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### FREE-Merging CoLA Score: 54.5. Result ID: 699997175698665ec40b2220. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### T-Switch CoLA Score: 53.12. Result ID: 699997175698665ec40b2222. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### TALL-Mask CoLA Score: 45.81. Result ID: 699997175698665ec40b2221. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### EMR-Merging CoLA Score: 39.96. Result ID: 699997175698665ec40b221f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### RegMean CoLA Score: 36.67. Result ID: 699997175698665ec40b2219. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Ties-Merging CoLA Score: 20.48. Result ID: 699997175698665ec40b221b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Task-Arithmetic CoLA Score: 18.78. Result ID: 699997175698665ec40b221a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Weight-Averaging CoLA Score: 13.96. Result ID: 699997175698665ec40b2218. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### DARE CoLA Score: 9.28. Result ID: 699997175698665ec40b221c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"RoBERTa"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.