# Agent Performance on AgentNoiseBench (Noisy setting tau2 1.0, test) Source: [Agent Performance on AgentNoiseBench (Noisy setting tau2 1.0, test)]() Collection index: [llms.txt]() Data updated: 2026-05-27T10:13:00.768Z Evaluation of agent performance under a specific noisy configuration (tau2 1.0) of the AgentNoiseBench dataset, reporting Avg@4 and Pass@4 scores for Retail, Airline, and Telecom domains. 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 | Agent Performance | | Dataset (including any stated split) | AgentNoiseBench Noisy setting tau2 1.0 (test) | | Metric | Retail Avg@4 Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 10 of 10 | | 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 - [Retail Avg@4 Score](): Higher is better; unit: PERCENT. - [Retail Pass@4 Score](): Higher is better; unit: PERCENT. - [Airline Avg@4 Score](): Higher is better; unit: PERCENT. - [Airline Pass@4 Score](): Higher is better; unit: PERCENT. - [Telecom Avg@4 Score](): Higher is better; unit: PERCENT. - [Telecom Pass@4 Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Retail Avg@4 Score | Result date | Report paper | | --- | --- | --- | --- | | NoisyAgent | 43.2 | Not recorded | 6a167ad3dbd7150506f80299 | | GRPO | 38.16 | Not recorded | 6a167ad3dbd7150506f80299 | | GSPO | 37.72 | Not recorded | 6a167ad3dbd7150506f80299 | | NoisyAgent | 36.4 | Not recorded | 6a167ad3dbd7150506f80299 | | DAPO | 36.18 | Not recorded | 6a167ad3dbd7150506f80299 | | GSPO | 31.8 | Not recorded | 6a167ad3dbd7150506f80299 | | Qwen3-32B | 31.14 | Not recorded | 6a167ad3dbd7150506f80299 | | GRPO | 30.48 | Not recorded | 6a167ad3dbd7150506f80299 | | DAPO | 29.39 | Not recorded | 6a167ad3dbd7150506f80299 | | Qwen3-8B | 24.12 | Not recorded | 6a167ad3dbd7150506f80299 | ### NoisyAgent Retail Avg@4 Score: 43.2. Result ID: 6a16c399fbc373592c1080bc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-32B","Training Method":"NoisyAgent"}. This record does not assert independent reproduction. ### GRPO Retail Avg@4 Score: 38.16. Result ID: 6a16c399fbc373592c1080b9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-32B","Training Method":"GRPO"}. This record does not assert independent reproduction. ### GSPO Retail Avg@4 Score: 37.72. Result ID: 6a16c399fbc373592c1080bb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-32B","Training Method":"GSPO"}. This record does not assert independent reproduction. ### NoisyAgent Retail Avg@4 Score: 36.4. Result ID: 6a16c399fbc373592c1080b7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Training Method":"NoisyAgent"}. This record does not assert independent reproduction. ### DAPO Retail Avg@4 Score: 36.18. Result ID: 6a16c399fbc373592c1080ba. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-32B","Training Method":"DAPO"}. This record does not assert independent reproduction. ### GSPO Retail Avg@4 Score: 31.8. Result ID: 6a16c399fbc373592c1080b6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Training Method":"GSPO"}. This record does not assert independent reproduction. ### Qwen3-32B Retail Avg@4 Score: 31.14. Result ID: 6a16c399fbc373592c1080b8. [Reporting paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-32B","Training Method":"Base"}. This record does not assert independent reproduction. ### GRPO Retail Avg@4 Score: 30.48. Result ID: 6a16c399fbc373592c1080b4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Training Method":"GRPO"}. This record does not assert independent reproduction. ### DAPO Retail Avg@4 Score: 29.39. Result ID: 6a16c399fbc373592c1080b5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Training Method":"DAPO"}. This record does not assert independent reproduction. ### Qwen3-8B Retail Avg@4 Score: 24.12. Result ID: 6a16c399fbc373592c1080b3. [Reporting paper]() Recorded evaluation configuration: {"Backbone Model":"Qwen3-8B","Training Method":"Base"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.