Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Comprehensive Evaluation on SeedBench (all)
Loading...
76.8
Score
1.7B K^4_{1.0}
37.4568
47.6709
57.885
68.0991
Mar 31, 2025
May 19, 2025
Jul 7, 2025
Aug 26, 2025
Oct 14, 2025
Dec 2, 2025
Jan 21, 2026
Score
Updated 3d ago
Evaluation Results
Method
Method
Links
Score
1.7B K^4_{1.0}
Size=18B-A3B, Sparsity...
2026.01
76.8
MetaGPT
merging_target=Qwen2-V...
2025.03
76.35
Ties-Merging
merging_target=Qwen2-V...
2025.03
76.29
AdaMMS
merging_target=Qwen2-V...
2025.03
76.2
Qwen2-VL (base)
model_status=Original...
2025.03
75.85
Task Arithmetic
merging_target=Qwen2-V...
2025.03
75.81
1.7B K^8_{0.5}
Size=18B-A3B, Sparsity...
2026.01
75.8
LLaVA-OneVision
model_status=Original...
2025.03
75.44
DARE-Linear
merging_target=Qwen2-V...
2025.03
74.32
Isaac 0.2
Size=2B
2026.01
74.2
DARE-Ties
merging_target=Qwen2-V...
2025.03
74.07
Task Arithmetic
Merging Protocol=Task...
2025.03
65.85
DARE-Linear
Merging Protocol=DARE-...
2025.03
65.07
AdaMMS
Merging Protocol=AdaMMS
2025.03
64.12
MetaGPT
Merging Protocol=MetaGPT
2025.03
61.29
CogVLM-chat-7B
Status=Base Model
2025.03
61.22
LLaVA-v1.5-7B
Status=Source Model
2025.03
60.52
DARE-Ties
Merging Protocol=DARE-...
2025.03
58.41
Ties-Merging
Merging Protocol=Ties-...
2025.03
38.97
Feedback
Search any
task
Search any
task