# General Chat Evaluation on Arena-Hard Source: [General Chat Evaluation on Arena-Hard]() Collection index: [llms.txt]() Data updated: 2026-03-06T08:28:42.755Z Evaluation of model performance on general chat capabilities using the base Arena-Hard dataset. 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 | General Chat Evaluation | | Dataset (including any stated split) | Arena-Hard | | Metric | Win Rate | | 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 - [Win Rate](): Higher is better; unit: PERCENT. - [SC](): Higher is better; unit: PERCENT. ## Recorded results | Method | Win Rate | Result date | Report paper | | --- | --- | --- | --- | | VRM-PPO | 84 | Not recorded | 69aa5f712d603e6de3e4182a | | DPO | 70.8 | Not recorded | 69aa5f712d603e6de3e4182a | | KTO | 69.5 | Not recorded | 69aa5f712d603e6de3e4182a | | VRM-PPO | 65.5 | Not recorded | 69aa5f712d603e6de3e4182a | | IPO | 65.5 | Not recorded | 69aa5f712d603e6de3e4182a | | WPO | 62 | Not recorded | 69aa5f712d603e6de3e4182a | | SIMPO | 61.9 | Not recorded | 69aa5f712d603e6de3e4182a | | SELECTIVE DPO | 60.4 | Not recorded | 69aa5f712d603e6de3e4182a | | PPO | 60.4 | Not recorded | 69aa5f712d603e6de3e4182a | | WPO | 56.2 | Not recorded | 69aa5f712d603e6de3e4182a | | PPO | 55.3 | Not recorded | 69aa5f712d603e6de3e4182a | | IPO | 54.6 | Not recorded | 69aa5f712d603e6de3e4182a | | SIMPO | 54.6 | Not recorded | 69aa5f712d603e6de3e4182a | | SELECTIVE DPO | 51.7 | Not recorded | 69aa5f712d603e6de3e4182a | | DPO | 49 | Not recorded | 69aa5f712d603e6de3e4182a | | KTO | 43.9 | Not recorded | 69aa5f712d603e6de3e4182a | ### VRM-PPO Win Rate: 84. Result ID: 69aa8f416dbb7075b9544a5f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### DPO Win Rate: 70.8. Result ID: 69aa8f416dbb7075b9544a58. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### KTO Win Rate: 69.5. Result ID: 69aa8f416dbb7075b9544a5a. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### VRM-PPO Win Rate: 65.5. Result ID: 69aa8f416dbb7075b9544a57. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### IPO Win Rate: 65.5. Result ID: 69aa8f416dbb7075b9544a59. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### WPO Win Rate: 62. Result ID: 69aa8f416dbb7075b9544a54. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### SIMPO Win Rate: 61.9. Result ID: 69aa8f416dbb7075b9544a5b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### SELECTIVE DPO Win Rate: 60.4. Result ID: 69aa8f416dbb7075b9544a5d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### PPO Win Rate: 60.4. Result ID: 69aa8f416dbb7075b9544a5e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### WPO Win Rate: 56.2. Result ID: 69aa8f416dbb7075b9544a5c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen 3-8B"}. This record does not assert independent reproduction. ### PPO Win Rate: 55.3. Result ID: 69aa8f416dbb7075b9544a56. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### IPO Win Rate: 54.6. Result ID: 69aa8f416dbb7075b9544a51. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### SIMPO Win Rate: 54.6. Result ID: 69aa8f416dbb7075b9544a53. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### SELECTIVE DPO Win Rate: 51.7. Result ID: 69aa8f416dbb7075b9544a55. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### DPO Win Rate: 49. Result ID: 69aa8f416dbb7075b9544a50. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### KTO Win Rate: 43.9. Result ID: 69aa8f416dbb7075b9544a52. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.