# Pairwise Preference Comparison on Pairwise Preference Comparisons 1.5B Scale (test) Source: [Pairwise Preference Comparison on Pairwise Preference Comparisons 1.5B Scale (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:36:18.319Z Evaluation of model performance using Pairwise Preference Comparison on the test set of the Pairwise Preference Comparisons 1.5B Scale dataset, measured by Average preference score. 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 | Pairwise Preference Comparison | | Dataset (including any stated split) | Pairwise Preference Comparisons 1.5B Scale (test) | | Metric | Avg Preference Score | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 30 of 30 | | 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 - [Avg Preference Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | Avg Preference Score | Result date | Report paper | | --- | --- | --- | --- | | STACKELBERGGDA FOLLOWER | 0.834 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA FOLLOWER | 0.825 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA FOLLOWER | 0.813 | Not recorded | 6944dc1f423ea42265729bf8 | | NASH-MD-PG | 0.812 | Not recorded | 6944dc1f423ea42265729bf8 | | NASH-MD-PG | 0.803 | Not recorded | 6944dc1f423ea42265729bf8 | | NASH-MD-PG | 0.791 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA LEADER | 0.729 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA LEADER | 0.707 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA LEADER | 0.69 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA FOLLOWER | 0.687 | Not recorded | 6944dc1f423ea42265729bf8 | | NASH-MD-PG | 0.623 | Not recorded | 6944dc1f423ea42265729bf8 | | RLOO | 0.621 | Not recorded | 6944dc1f423ea42265729bf8 | | RLOO | 0.599 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-1.5B | 0.521 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA FOLLOWER | 0.511 | Not recorded | 6944dc1f423ea42265729bf8 | | NASH-MD-PG | 0.489 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-0.5B | 0.479 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-1.5B | 0.401 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-0.5B | 0.379 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA LEADER | 0.377 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA LEADER | 0.313 | Not recorded | 6944dc1f423ea42265729bf8 | | RLOO | 0.31 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-1.5B | 0.293 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-0.5B | 0.271 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-1.5B | 0.209 | Not recorded | 6944dc1f423ea42265729bf8 | | RLOO | 0.197 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-0.5B | 0.188 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-1.5B | 0.187 | Not recorded | 6944dc1f423ea42265729bf8 | | RLOO | 0.175 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-0.5B | 0.166 | Not recorded | 6944dc1f423ea42265729bf8 | ### STACKELBERGGDA FOLLOWER Avg Preference Score: 0.834. Result ID: 699952975698665ec4bd68ba. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-0.5B"}. This record does not assert independent reproduction. ### STACKELBERGGDA FOLLOWER Avg Preference Score: 0.825. Result ID: 699952975698665ec4bd68bc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"RLOO"}. This record does not assert independent reproduction. ### STACKELBERGGDA FOLLOWER Avg Preference Score: 0.813. Result ID: 699952975698665ec4bd68bb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-1.5B"}. This record does not assert independent reproduction. ### NASH-MD-PG Avg Preference Score: 0.812. Result ID: 699952975698665ec4bd68b0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-0.5B"}. This record does not assert independent reproduction. ### NASH-MD-PG Avg Preference Score: 0.803. Result ID: 699952975698665ec4bd68b2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"RLOO"}. This record does not assert independent reproduction. ### NASH-MD-PG Avg Preference Score: 0.791. Result ID: 699952975698665ec4bd68b1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-1.5B"}. This record does not assert independent reproduction. ### STACKELBERGGDA LEADER Avg Preference Score: 0.729. Result ID: 699952975698665ec4bd68b5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-0.5B"}. This record does not assert independent reproduction. ### STACKELBERGGDA LEADER Avg Preference Score: 0.707. Result ID: 699952975698665ec4bd68b6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-1.5B"}. This record does not assert independent reproduction. ### STACKELBERGGDA LEADER Avg Preference Score: 0.69. Result ID: 699952975698665ec4bd68b7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"RLOO"}. This record does not assert independent reproduction. ### STACKELBERGGDA FOLLOWER Avg Preference Score: 0.687. Result ID: 699952975698665ec4bd68be. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"STACKELBERGGDA LEADER"}. This record does not assert independent reproduction. ### NASH-MD-PG Avg Preference Score: 0.623. Result ID: 699952975698665ec4bd68b3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"STACKELBERGGDA LEADER"}. This record does not assert independent reproduction. ### RLOO Avg Preference Score: 0.621. Result ID: 699952975698665ec4bd68ab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-0.5B"}. This record does not assert independent reproduction. ### RLOO Avg Preference Score: 0.599. Result ID: 699952975698665ec4bd68ac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"QWEN2.5-1.5B"}. This record does not assert independent reproduction. ### QWEN2.5-1.5B Avg Preference Score: 0.521. Result ID: 699952975698665ec4bd68a6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"QWEN2.5-0.5B"}. This record does not assert independent reproduction. ### STACKELBERGGDA FOLLOWER Avg Preference Score: 0.511. Result ID: 699952975698665ec4bd68bd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"NASH-MD-PG"}. This record does not assert independent reproduction. ### NASH-MD-PG Avg Preference Score: 0.489. Result ID: 699952975698665ec4bd68b4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"STACKELBERGGDA FOLLOWER"}. This record does not assert independent reproduction. ### QWEN2.5-0.5B Avg Preference Score: 0.479. Result ID: 699952975698665ec4bd68a1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"QWEN2.5-1.5B"}. This record does not assert independent reproduction. ### QWEN2.5-1.5B Avg Preference Score: 0.401. Result ID: 699952975698665ec4bd68a7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"RLOO"}. This record does not assert independent reproduction. ### QWEN2.5-0.5B Avg Preference Score: 0.379. Result ID: 699952975698665ec4bd68a2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"RLOO"}. This record does not assert independent reproduction. ### STACKELBERGGDA LEADER Avg Preference Score: 0.377. Result ID: 699952975698665ec4bd68b8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"NASH-MD-PG"}. This record does not assert independent reproduction. ### STACKELBERGGDA LEADER Avg Preference Score: 0.313. Result ID: 699952975698665ec4bd68b9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"STACKELBERGGDA FOLLOWER"}. This record does not assert independent reproduction. ### RLOO Avg Preference Score: 0.31. Result ID: 699952975698665ec4bd68ae. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"STACKELBERGGDA LEADER"}. This record does not assert independent reproduction. ### QWEN2.5-1.5B Avg Preference Score: 0.293. Result ID: 699952975698665ec4bd68a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"STACKELBERGGDA LEADER"}. This record does not assert independent reproduction. ### QWEN2.5-0.5B Avg Preference Score: 0.271. Result ID: 699952975698665ec4bd68a4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"STACKELBERGGDA LEADER"}. This record does not assert independent reproduction. ### QWEN2.5-1.5B Avg Preference Score: 0.209. Result ID: 699952975698665ec4bd68a8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"NASH-MD-PG"}. This record does not assert independent reproduction. ### RLOO Avg Preference Score: 0.197. Result ID: 699952975698665ec4bd68ad. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"NASH-MD-PG"}. This record does not assert independent reproduction. ### QWEN2.5-0.5B Avg Preference Score: 0.188. Result ID: 699952975698665ec4bd68a3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"NASH-MD-PG"}. This record does not assert independent reproduction. ### QWEN2.5-1.5B Avg Preference Score: 0.187. Result ID: 699952975698665ec4bd68aa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"STACKELBERGGDA FOLLOWER"}. This record does not assert independent reproduction. ### RLOO Avg Preference Score: 0.175. Result ID: 699952975698665ec4bd68af. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"QWEN2.5-1.5B","Opponent Model":"STACKELBERGGDA FOLLOWER"}. This record does not assert independent reproduction. ### QWEN2.5-0.5B Avg Preference Score: 0.166. Result ID: 699952975698665ec4bd68a5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Opponent Model":"STACKELBERGGDA FOLLOWER"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.