# Multi-constraint steering on RECAST-5 Source: [Multi-constraint steering on RECAST-5]() Collection index: [llms.txt]() Data updated: 2026-05-29T10:53:48.910Z Evaluation of model performance on the RECAST-5 dataset focusing on multi-constraint steering capabilities, measured by the R5 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-5 | | Metric | R5 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 - [R5 Score](): Higher is better; unit: SCALAR. ## Recorded results | Method | R5 Score | Result date | Report paper | | --- | --- | --- | --- | | Original | 21.49 | Not recorded | 6a191de3dbd7150506f82bf4 | | UNISTEER | 21.49 | Not recorded | 6a191de3dbd7150506f82bf4 | | RepE | 21.29 | Not recorded | 6a191de3dbd7150506f82bf4 | | CAA | 20.68 | Not recorded | 6a191de3dbd7150506f82bf4 | | LoReFT | 20.06 | Not recorded | 6a191de3dbd7150506f82bf4 | | ODESteer | 15.06 | Not recorded | 6a191de3dbd7150506f82bf4 | | UNISTEER | 14.66 | Not recorded | 6a191de3dbd7150506f82bf4 | | RepE | 14.46 | Not recorded | 6a191de3dbd7150506f82bf4 | | ODESteer | 13.66 | Not recorded | 6a191de3dbd7150506f82bf4 | | CAA | 13.65 | Not recorded | 6a191de3dbd7150506f82bf4 | | Original | 12.85 | Not recorded | 6a191de3dbd7150506f82bf4 | | RepE | 12.85 | Not recorded | 6a191de3dbd7150506f82bf4 | | UNISTEER | 12.85 | Not recorded | 6a191de3dbd7150506f82bf4 | | LoReFT | 12.65 | Not recorded | 6a191de3dbd7150506f82bf4 | | CAA | 11.24 | Not recorded | 6a191de3dbd7150506f82bf4 | | Original | 11.04 | Not recorded | 6a191de3dbd7150506f82bf4 | | LoReFT | 10.44 | Not recorded | 6a191de3dbd7150506f82bf4 | | ODESteer | 9.61 | Not recorded | 6a191de3dbd7150506f82bf4 | ### Original R5 Score: 21.49. Result ID: 6a196fdcfbc373592c148ca6. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### UNISTEER R5 Score: 21.49. Result ID: 6a196fdcfbc373592c148cab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### RepE R5 Score: 21.29. Result ID: 6a196fdcfbc373592c148ca8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### CAA R5 Score: 20.68. Result ID: 6a196fdcfbc373592c148ca7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### LoReFT R5 Score: 20.06. Result ID: 6a196fdcfbc373592c148ca9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### ODESteer R5 Score: 15.06. Result ID: 6a196fdcfbc373592c148caa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-7B"}. This record does not assert independent reproduction. ### UNISTEER R5 Score: 14.66. Result ID: 6a196fdcfbc373592c148c9f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### RepE R5 Score: 14.46. Result ID: 6a196fdcfbc373592c148c9c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### ODESteer R5 Score: 13.66. Result ID: 6a196fdcfbc373592c148ca4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### CAA R5 Score: 13.65. Result ID: 6a196fdcfbc373592c148c9b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### Original R5 Score: 12.85. Result ID: 6a196fdcfbc373592c148c9a. [Reporting paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### RepE R5 Score: 12.85. Result ID: 6a196fdcfbc373592c148ca2. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### UNISTEER R5 Score: 12.85. Result ID: 6a196fdcfbc373592c148ca5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### LoReFT R5 Score: 12.65. Result ID: 6a196fdcfbc373592c148c9d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Llama-3.2-1B"}. This record does not assert independent reproduction. ### CAA R5 Score: 11.24. Result ID: 6a196fdcfbc373592c148ca1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### Original R5 Score: 11.04. Result ID: 6a196fdcfbc373592c148ca0. [Reporting paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### LoReFT R5 Score: 10.44. Result ID: 6a196fdcfbc373592c148ca3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Qwen2.5-1.5B"}. This record does not assert independent reproduction. ### ODESteer R5 Score: 9.61. Result ID: 6a196fdcfbc373592c148c9e. [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.