Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-modal Needle-In-A-Haystack on MMNeedle COCO2014-based (test)
Loading...
42.5
Exact Accuracy (8x8)
LLaVA-Llama-3-8B + BEFT
-1.7
9.775
21.25
32.725
Apr 24, 2026
Exact Accuracy (8x8)
Exact Accuracy (1x1)
Exact Accuracy (2x2)
Exact Accuracy (4x4)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Exact Accuracy (8x8)
Exact Accuracy (1x1)
Exact Accuracy (2x2)
Exact Accuracy (4x4)
LLaVA-Llama-3-8B + BEFT
M (Number of image pan...
2026.04
42.5
-
-
-
GPT-4o
M (Number of image pan...
2026.04
1
97
81.8
26.9
Claude 3 Opus
M (Number of image pan...
2026.04
0
66.9
4.6
0.4
LLaVA-Llama-3-8B
M (Number of image pan...
2026.04
0
0
0
0
LLaVA-Llama-3-8B + BEFT
M (Number of image pan...
2026.04
0
97.1
86.8
41.4
Feedback
Search any
task
Search any
task