# Multi-constraint steering on RECAST-10 Source: [Multi-constraint steering on RECAST-10]() Collection index: [llms.txt]() Data updated: 2026-05-29T10:53:48.937Z Evaluation of model performance on the RECAST-10 dataset focusing on multi-constraint steering capabilities, measured by the R10 metric. 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 | Multi-constraint steering | | Dataset (including any stated split) | RECAST-10 | | Metric | R10 Score | | Unit | SCALAR | | 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 - [R10 Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | R10 Score | Result date | Report paper | | --- | --- | --- | --- | | UNISTEER | 13.05 | Not recorded | 6a191de3dbd7150506f82bf4 | | RepE | 11.45 | Not recorded | 6a191de3dbd7150506f82bf4 | | Original | 10.44 | Not recorded | 6a191de3dbd7150506f82bf4 | | CAA | 10.04 | Not recorded | 6a191de3dbd7150506f82bf4 | | ODESteer | 9.04 | Not recorded | 6a191de3dbd7150506f82bf4 | | LoReFT | 8.84 | Not recorded | 6a191de3dbd7150506f82bf4 | | UNISTEER | 6.83 | Not recorded | 6a191de3dbd7150506f82bf4 | | UNISTEER | 6.63 | Not recorded | 6a191de3dbd7150506f82bf4 | | CAA | 5.82 | Not recorded | 6a191de3dbd7150506f82bf4 | | RepE | 5.82 | Not recorded | 6a191de3dbd7150506f82bf4 | | Original | 5.62 | Not recorded | 6a191de3dbd7150506f82bf4 | | LoReFT | 5.22 | Not recorded | 6a191de3dbd7150506f82bf4 | | CAA | 5.22 | Not recorded | 6a191de3dbd7150506f82bf4 | | ODESteer | 5.04 | Not recorded | 6a191de3dbd7150506f82bf4 | | Original | 5.02 | Not recorded | 6a191de3dbd7150506f82bf4 | | RepE | 4.82 | Not recorded | 6a191de3dbd7150506f82bf4 | | LoReFT | 4.81 | Not recorded | 6a191de3dbd7150506f82bf4 | | ODESteer | 3.61 | Not recorded | 6a191de3dbd7150506f82bf4 | ### UNISTEER R10 Score: 13.05. Result ID: 6a196fdcfbc373592c148cbd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### RepE R10 Score: 11.45. Result ID: 6a196fdcfbc373592c148cba. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### Original R10 Score: 10.44. Result ID: 6a196fdcfbc373592c148cb8. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### CAA R10 Score: 10.04. Result ID: 6a196fdcfbc373592c148cb9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### ODESteer R10 Score: 9.04. Result ID: 6a196fdcfbc373592c148cbc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### LoReFT R10 Score: 8.84. Result ID: 6a196fdcfbc373592c148cbb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### UNISTEER R10 Score: 6.83. Result ID: 6a196fdcfbc373592c148cb1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### UNISTEER R10 Score: 6.63. Result ID: 6a196fdcfbc373592c148cb7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### CAA R10 Score: 5.82. Result ID: 6a196fdcfbc373592c148cad. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### RepE R10 Score: 5.82. Result ID: 6a196fdcfbc373592c148cae. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### Original R10 Score: 5.62. Result ID: 6a196fdcfbc373592c148cac. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### LoReFT R10 Score: 5.22. Result ID: 6a196fdcfbc373592c148caf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### CAA R10 Score: 5.22. Result ID: 6a196fdcfbc373592c148cb3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### ODESteer R10 Score: 5.04. Result ID: 6a196fdcfbc373592c148cb6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### Original R10 Score: 5.02. Result ID: 6a196fdcfbc373592c148cb2. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### RepE R10 Score: 4.82. Result ID: 6a196fdcfbc373592c148cb4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### LoReFT R10 Score: 4.81. Result ID: 6a196fdcfbc373592c148cb5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### ODESteer R10 Score: 3.61. Result ID: 6a196fdcfbc373592c148cb0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.