List your company

Closed-ended Task T-1 on DanmakuTPP-QA

27.33AccuracyMM-TPP
-0.756.5413.8321.12Feb 1, 2026

Evaluation results

MethodLinks
MM-TPPModel Category=Finetuned2026.0227.33
Qwen2.5-VL-3BModel Category=Finetuned2026.0227
Qwen2.5-32B-InstructModel Category=LLMs2026.0225.33
DeepSeek-V3Model Category=LLMs2026.0225
Qwen2.5-VL-7BModel Category=MLLMs2026.029.67
Qwen3-235B-A22BModel Category=LLMs2026.028.67
Qwen2.5-VL-32BModel Category=MLLMs2026.028
Qwen3-8BModel Category=LLMs2026.026.67
Llama-3.3-70B-InstructModel Category=LLMs2026.021.67
Qwen2.5-72B-InstructModel Category=LLMs2026.020.67
Qwen3-30B-A3BModel Category=LLMs2026.020.67
Qwen2.5-7B-InstructModel Category=LLMs2026.020.33
Qwen2.5-VL-72BModel Category=MLLMs2026.020.33
Gemma3-27BModel Category=MLLMs2026.020.33