Overall performance on CoP-QA-F
97.6AC ScoreTalk2DM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Talk2DMLLM Backbone=Gemma3:27B2026.02 | 97.6 | 93.09 | |
| Talk2DMLLM Backbone=GPT-oss:20B2026.02 | 96.81 | 94.45 | |
| Talk2DMLLM Backbone=Qwen3:30B2026.02 | 96.55 | 95.15 | |
| Talk2DMLLM Backbone=Magistral:24B2026.02 | 90.81 | 71.78 | |
| Talk2DMLLM Backbone=Deepseek-r1:32B2026.02 | 86.85 | 85.17 | |
| Talk2DMLLM Backbone=Llama3.1:8B2026.02 | 79.16 | 71.46 |