List your company

Proactive Assistance on Pro2Bench EgoProactive

68G-Mean F1Qwen3.6-VL-27B
-2.7215.643452.36Jun 3, 2026

Evaluation results

MethodLinks
Qwen3.6-VL-27BFT=true, PWR=true, Architecture=split duplex+planner2026.066844
Claude Opus 4.6Architecture=monolithic, Mode=zero-shot2026.065743
Qwen3-VL-235BArchitecture=monolithic2026.065733
Llama 4-MaverickFT=true, PWR=true, Architecture=split duplex+planner2026.065743
Llama 4-MaverickArchitecture=monolithic2026.064934
Llama 4-MaverickFT=true, Architecture=monolithic2026.064940
Gemini 3.1 ProArchitecture=monolithic, Mode=zero-shot2026.064638
Qwen3.6-VL-27BArchitecture=monolithic2026.062824
Qwen3.6-VL-27BFT=true, Architecture=monolithic2026.062246
GPT-5.2Architecture=monolithic, Mode=zero-shot2026.060—