# Evidence classification (Task A) on TREC BioGen 2025 Source: [Evidence classification (Task A) on TREC BioGen 2025]() Collection index: [llms.txt]() Data updated: 2026-03-19T09:36:23.377Z Evaluation of evidence classification performance (Task A) on the TREC BioGen 2025 dataset, utilizing supporting and contributing precision, recall, and F1 metrics. 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 | Evidence classification (Task A) | | Dataset (including any stated split) | TREC BioGen 2025 | | Metric | Supporting Precision | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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 - [Supporting Precision](): Higher is better; unit: PERCENT. - [Supporting Recall](): Higher is better; unit: PERCENT. - [Supporting F1 Score](): Higher is better; unit: PERCENT. - [Contributing Precision](): Higher is better; unit: PERCENT. - [Contributing Recall](): Higher is better; unit: PERCENT. - [Contributing F1 Score](): Higher is better; unit: PERCENT. - [Strict Support Precision](): Higher is better; unit: PERCENT. - [Strict Support Recall](): Higher is better; unit: PERCENT. - [Strict Contradiction Precision](): Higher is better; unit: PERCENT. - [Strict Contradiction Recall](): Higher is better; unit: PERCENT. - [Relaxed Support Precision](): Higher is better; unit: PERCENT. - [Relaxed Support Recall](): Higher is better; unit: PERCENT. - [Relaxed Contradiction Precision](): Higher is better; unit: PERCENT. - [Relaxed Contradiction Recall](): Higher is better; unit: PERCENT. ## Recorded results | Method | Supporting Precision | Result date | Report paper | | --- | --- | --- | --- | | LLM\_NLI\_BM25 | 67.18 | Not recorded | 69bb83260636e5bf4afc1bc1 | | run6\_A | 66.92 | Not recorded | 69bb83260636e5bf4afc1bc1 | | LLM\_BM25 | 66.75 | Not recorded | 69bb83260636e5bf4afc1bc1 | | task\_a | 56.7 | Not recorded | 69bb83260636e5bf4afc1bc1 | | run4 | 52.92 | Not recorded | 69bb83260636e5bf4afc1bc1 | | run2 | 52.75 | Not recorded | 69bb83260636e5bf4afc1bc1 | | scifive-ft | 52.58 | Not recorded | 69bb83260636e5bf4afc1bc1 | | task-a-1 | 52.41 | Not recorded | 69bb83260636e5bf4afc1bc1 | | TEST | 51.03 | Not recorded | 69bb83260636e5bf4afc1bc1 | | emotional | 50.6 | Not recorded | 69bb83260636e5bf4afc1bc1 | ### LLM\_NLI\_BM25 Supporting Precision: 67.18. Result ID: 69bbc39147f34ce924a60498. [Reporting paper]() Recorded evaluation configuration: {"Team":"CLaC"}. This record does not assert independent reproduction. ### run6\_A Supporting Precision: 66.92. Result ID: 69bbc39147f34ce924a60495. [Reporting paper]() Recorded evaluation configuration: {"Team":"InfoLab"}. This record does not assert independent reproduction. ### LLM\_BM25 Supporting Precision: 66.75. Result ID: 69bbc39147f34ce924a60499. [Reporting paper]() Recorded evaluation configuration: {"Team":"CLaC"}. This record does not assert independent reproduction. ### task\_a Supporting Precision: 56.7. Result ID: 69bbc39147f34ce924a60497. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Team":"GEHC-HTIC"}. This record does not assert independent reproduction. ### run4 Supporting Precision: 52.92. Result ID: 69bbc39147f34ce924a60496. [Reporting paper]() Recorded evaluation configuration: {"Team":"InfoLab"}. This record does not assert independent reproduction. ### run2 Supporting Precision: 52.75. Result ID: 69bbc39147f34ce924a60494. [Reporting paper]() Recorded evaluation configuration: {"Team":"InfoLab"}. This record does not assert independent reproduction. ### scifive-ft Supporting Precision: 52.58. Result ID: 69bbc39147f34ce924a6049a. [Reporting paper]() Recorded evaluation configuration: {"Team":"polito"}. This record does not assert independent reproduction. ### task-a-1 Supporting Precision: 52.41. Result ID: 69bbc39147f34ce924a6049b. [Reporting paper]() Recorded evaluation configuration: {"Team":"SIB"}. This record does not assert independent reproduction. ### TEST Supporting Precision: 51.03. Result ID: 69bbc39147f34ce924a6049c. [Reporting paper]() Recorded evaluation configuration: {"Team":"Baseline"}. This record does not assert independent reproduction. ### emotional Supporting Precision: 50.6. Result ID: 69bbc39147f34ce924a6049d. [Reporting paper]() Recorded evaluation configuration: {"Team":"dal"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.