List your company

Scientific Table & Figure Analysis on ANABENCH (test)

61.06Cosine SimilarityANAGENT
52.06454.399556.73559.0705Feb 10, 2026

Evaluation results

MethodLinks
ANAGENTEvaluation Protocol=One-Shot, Backbone=Gemini-2.5-Flash, Size=-2026.0261.0661.3420.5217.44.0611.4747.6410.9829.31
ANAGENTEvaluation Protocol=One-Shot, Backbone=GPT-4.1-mini, Size=-2026.0260.8763.2824.2620.655.7312.5549.4712.9831.22
ANAGENTEvaluation Protocol=One-Shot, Backbone=Qwen3-VL, Size=8B2026.0260.5562.2724.6520.065.9212.9549.1512.9831.07
ANAGENTEvaluation Protocol=One-Shot, Backbone=InternVL-3.5, Size=8B2026.0260.2662.1223.1819.144.5612.9748.5212.2230.37
ANAGENTEvaluation Protocol=One-Shot, Backbone=Qwen2.5-VL, Size=7B2026.0260.246123.4119.414.9812.4748.2212.2930.25
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=GPT-4.1-mini, Size=-2026.0259.9461.6322.7518.194.8112.2648.1111.7529.93
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=Qwen3-VL, Size=8B2026.0259.7661.5323.0717.754.9812.248.1211.6429.88
ANAGENTEvaluation Protocol=One-Shot, Backbone=Qwen3-VL, Size=4B2026.0259.6460.6122.4218.054.0311.5147.5511.229.38
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=InternVL-3.5, Size=8B2026.0259.4661.2522.59173.8811.6847.7710.8529.31
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=Qwen3-VL, Size=4B2026.0259.4160.2121.2316.273.911.3346.9510.528.73
ANAGENTEvaluation Protocol=One-Shot, Backbone=InternVL-3.5, Size=4B2026.0259.1160.5222.618.043.8211.547.4111.1229.27
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=Qwen2.5-VL, Size=7B2026.0258.9160.4121.5917.474.1111.8546.9711.1429.06
ANAGENTEvaluation Protocol=One-Shot, Backbone=Qwen2.5-VL, Size=3B2026.0258.8960.722.1918.413.9911.5447.2611.3129.29
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=InternVL-3.5, Size=4B2026.0258.2659.8621.2116.13.2911.1146.4410.1728.31
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=Qwen2.5-VL, Size=3B2026.0257.560.0121.0317.343.8711.5346.1810.9128.55
GPT-4.1-miniEvaluation Protocol=Baseline, Size=-2026.0256.3459.7419.4716.743.3911.4945.1810.5427.86
Qwen3-VLEvaluation Protocol=Baseline, Size=8B2026.0255.9459.1119.1617.063.0210.3944.7310.1627.44
InternVL-3.5Evaluation Protocol=Baseline, Size=8B2026.0255.7359.119.316.82.8610.2844.719.9827.34
Qwen2.5-VLEvaluation Protocol=Baseline, Size=7B2026.0255.6559.6618.916.42.9810.3844.749.9827.31
ANAGENTEvaluation Protocol=Zero-Shot, Backbone=Gemini-2.5-Flash, Size=-2026.0255.659.3719.416.043.1511.144.7910.0927.44
Qwen3-VLEvaluation Protocol=Baseline, Size=4B2026.0255.4158.1518.4115.772.7710.0643.999.5326.76
Qwen2.5-VLEvaluation Protocol=Baseline, Size=3B2026.0254.7458.4917.8215.892.5610.0243.689.4926.59
InternVL-3.5Evaluation Protocol=Baseline, Size=4B2026.0254.3858.1918.7615.672.669.843.789.3726.58
Gemini-2.5-FlashEvaluation Protocol=Baseline, Size=-2026.0252.4155.9919.0114.92.769.9542.479.225.84