# Test-Time Scaling Selection on RouterBench Cost-Sensitive Source: [Test-Time Scaling Selection on RouterBench Cost-Sensitive]() Collection index: [llms.txt]() Data updated: 2026-06-01T08:34:35.440Z Evaluation of model performance focusing on selecting the appropriate scaling configuration during test time, using the cost-sensitive variant of the RouterBench dataset. Metrics include Reward, Accuracy, and Cost. 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 | Test-Time Scaling Selection | | Dataset (including any stated split) | RouterBench Cost-Sensitive | | Metric | Reward | | Unit | SCALAR | | 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 - [Reward](): Higher is better; unit: SCALAR. - [Accuracy](): Higher is better; unit: PERCENT. - [Cost](): Lower is better; unit: SCALAR. ## Recorded results | Method | Reward | Result date | Report paper | | --- | --- | --- | --- | | Oracle | 0.8337 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | Oracle | 0.8297 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | UNISCALE | 0.7535 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | Greedy | 0.7184 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | UNISCALE | 0.7079 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | TS | 0.7022 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | MLP | 0.7006 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | NeuralUCB | 0.6984 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | k-NN | 0.6966 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | Random | 0.6937 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | k-NN | 0.659 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | MLP | 0.6301 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | NeuralUCB | 0.588 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | Random | 0.5731 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | Greedy | 0.5589 | Not recorded | 6a1d11e0dbd7150506f83dd0 | | TS | 0.5549 | Not recorded | 6a1d11e0dbd7150506f83dd0 | ### Oracle Reward: 0.8337. Result ID: 6a1d42ecfbc373592c167300. [Reporting paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### Oracle Reward: 0.8297. Result ID: 6a1d42ecfbc373592c1672f0. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### UNISCALE Reward: 0.7535. Result ID: 6a1d42ecfbc373592c1672ef. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### Greedy Reward: 0.7184. Result ID: 6a1d42ecfbc373592c1672ea. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### UNISCALE Reward: 0.7079. Result ID: 6a1d42ecfbc373592c1672ff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### TS Reward: 0.7022. Result ID: 6a1d42ecfbc373592c1672ee. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### MLP Reward: 0.7006. Result ID: 6a1d42ecfbc373592c1672eb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### NeuralUCB Reward: 0.6984. Result ID: 6a1d42ecfbc373592c1672ed. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### k-NN Reward: 0.6966. Result ID: 6a1d42ecfbc373592c1672ec. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### Random Reward: 0.6937. Result ID: 6a1d42ecfbc373592c1672e9. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-0.6B","Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### k-NN Reward: 0.659. Result ID: 6a1d42ecfbc373592c1672fc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### MLP Reward: 0.6301. Result ID: 6a1d42ecfbc373592c1672fb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### NeuralUCB Reward: 0.588. Result ID: 6a1d42ecfbc373592c1672fd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### Random Reward: 0.5731. Result ID: 6a1d42ecfbc373592c1672f9. [Reporting paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### Greedy Reward: 0.5589. Result ID: 6a1d42ecfbc373592c1672fa. [Reporting paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. ### TS Reward: 0.5549. Result ID: 6a1d42ecfbc373592c1672fe. [Reporting paper]() Recorded evaluation configuration: {"Number of seeds":"5","Iterations":"160","Warm-up phase":"50-step"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.