Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Report Generation on DTU (test)
Loading...
41
BLEU-4
Eyes + Bridge + QLoRA + RAFT
5.64
14.82
24
33.18
May 26, 2026
BLEU-4
HR
Expert Score
Updated 7d ago
Evaluation Results
Method
Method
Links
BLEU-4
HR
Expert Score
Eyes + Bridge + QLoRA + RAFT
Visual grounding=Text...
2026.05
41
4
8.6
Eyes + Bridge + QLoRA
Visual grounding=Text...
2026.05
36
18
7.4
Eyes + Bridge + DeepSeek-V3
Visual grounding=Text...
2026.05
19
29
5.9
Eyes + Bridge + Qwen base
Visual grounding=Text...
2026.05
14
38
4.6
Eyes + LLM, no Bridge
Visual grounding=Class...
2026.05
12
49
4.7
Prompt CoT (DeepSeek-V3, no Bridge)
Visual grounding=Class...
2026.05
9
61
3.8
Zero-shot VLM (GPT-4V)
Visual grounding=Full...
2026.05
7
65
3.3
Feedback
Search any
task
Search any
task