ResearchBenchmarksClassification Querying on CoP-QA-FFollow100AC ScoreTalk2DM87.395290.667693.9497.2124Feb 12, 2026Evaluation ResultsMethodMethodLinksAC ScoreAQ ScoreTalk2DMLLM Backbone=Qwen3:30BLLM Backbone=Qwen3:30B2026.02100100Talk2DMLLM Backbone=GPT-oss:20BLLM Backbone=GPT-oss:20B2026.0210097.47Talk2DMLLM Backbone=Gemma3:27BLLM Backbone=Gemma3:27B2026.0210084.11Talk2DMLLM Backbone=Llama3.1:8BLLM Backbone=Llama3.1:8B2026.0299.891.65Talk2DMLLM Backbone=Magistral...LLM Backbone=Magistral:24B2026.0298.9875.36Talk2DMLLM Backbone=Deepseek-...LLM Backbone=Deepseek-r1:32B2026.0287.8887.88