# Conversational Agent Interaction on Du-Interact Golden Benchmark Source: [Conversational Agent Interaction on Du-Interact Golden Benchmark]() Collection index: [llms.txt]() Data updated: 2026-03-23T10:24:04.876Z Evaluation of Conversational Agent Interaction performance on the Du-Interact Golden Benchmark, utilizing metrics such as Success Rate, Fidelity, Persona, Empathy, and average latency. 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 | Conversational Agent Interaction | | Dataset (including any stated split) | Du-Interact Golden Benchmark | | Metric | Dispatch Rate (P\_disp) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 9 of 9 | | 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 - [Dispatch Rate (P\_disp)](): Higher is better; unit: PERCENT. - [Success Rate (SR)](): Higher is better; unit: PERCENT. - [Fidelity](): Higher is better; unit: PERCENT. - [Persona Score (1-5)](): Higher is better; unit: SCALAR. - [Empathy Score (1-5)](): Higher is better; unit: SCALAR. - [Average Latency (ms)](): Lower is better; unit: TIME. ## Recorded results | Method | Dispatch Rate (P\_disp) | Result date | Report paper | | --- | --- | --- | --- | | DuCCAE-V3 | 82.5 | Not recorded | 69c0c9000636e5bf4afc3b29 | | DuCCAE-V2 | 76.2 | Not recorded | 69c0c9000636e5bf4afc3b29 | | Llama-3.3-70B-Instruct | 75.5 | Not recorded | 69c0c9000636e5bf4afc3b29 | | Qwen2.5-32B-Instruct | 72.1 | Not recorded | 69c0c9000636e5bf4afc3b29 | | DuCCAE-V1 | 68.8 | Not recorded | 69c0c9000636e5bf4afc3b29 | | Qwen2.5-7B-Instruct | 61.5 | Not recorded | 69c0c9000636e5bf4afc3b29 | | Llama-3.2-11B-Instruct | 58.2 | Not recorded | 69c0c9000636e5bf4afc3b29 | | Qwen2.5-3B-Instruct | 52.4 | Not recorded | 69c0c9000636e5bf4afc3b29 | | Llama-3.2-3B-Instruct | 51.8 | Not recorded | 69c0c9000636e5bf4afc3b29 | ### DuCCAE-V3 Dispatch Rate (P\_disp): 82.5. Result ID: 69c1136147f34ce924a91be1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setup":"Evolution II"}. This record does not assert independent reproduction. ### DuCCAE-V2 Dispatch Rate (P\_disp): 76.2. Result ID: 69c1136147f34ce924a91be0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setup":"Evolution I"}. This record does not assert independent reproduction. ### Llama-3.3-70B-Instruct Dispatch Rate (P\_disp): 75.5. Result ID: 69c1136147f34ce924a91bde. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Tier":"Large Scale Class (30B-70B)","Setup":"Zero-shot"}. This record does not assert independent reproduction. ### Qwen2.5-32B-Instruct Dispatch Rate (P\_disp): 72.1. Result ID: 69c1136147f34ce924a91bdd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Tier":"Large Scale Class (30B-70B)","Setup":"Zero-shot"}. This record does not assert independent reproduction. ### DuCCAE-V1 Dispatch Rate (P\_disp): 68.8. Result ID: 69c1136147f34ce924a91bdf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Setup":"Cold Start"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct Dispatch Rate (P\_disp): 61.5. Result ID: 69c1136147f34ce924a91bdb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Tier":"Server Class (7B-11B)","Setup":"Zero-shot"}. This record does not assert independent reproduction. ### Llama-3.2-11B-Instruct Dispatch Rate (P\_disp): 58.2. Result ID: 69c1136147f34ce924a91bdc. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Tier":"Server Class (7B-11B)","Setup":"Zero-shot"}. This record does not assert independent reproduction. ### Qwen2.5-3B-Instruct Dispatch Rate (P\_disp): 52.4. Result ID: 69c1136147f34ce924a91bd9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Tier":"Edge Class (3B)","Setup":"Zero-shot"}. This record does not assert independent reproduction. ### Llama-3.2-3B-Instruct Dispatch Rate (P\_disp): 51.8. Result ID: 69c1136147f34ce924a91bda. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Tier":"Edge Class (3B)","Setup":"Zero-shot"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.