List your company

Short-answer Question Answering on HotPotQA (EM, F1)

59.6F1AGAR
17.16828.18439.250.216May 10, 2026May 15, 2026May 20, 2026May 26, 2026May 31, 2026Jun 5, 2026Jun 11, 2026

Evaluation results

MethodLinks
AGAR2026.0659.6—
CGR2026.0657.1—
Random2026.0656.9—
Plain VQA2026.0656.4—
Uniform2026.0653.4—
VAR2026.0652.9—
VEA-Br2026.0651—
VEA-Con2026.0650.5—
AGLA2026.0649.2—
AGARBackbone=GLM-4.1V-9B-Thinking2026.0648—
AGLABackbone=GLM-4.1V-9B-Thinking2026.0643.2—
UniformBackbone=GLM-4.1V-9B-Thinking2026.0641.4—
RandomBackbone=GLM-4.1V-9B-Thinking2026.0641—
Plain VQABackbone=GLM-4.1V-9B-Thinking2026.0640.9—
VEA-ConBackbone=GLM-4.1V-9B-Thinking2026.0640.7—
APCDBackbone=Llama-2-13b-chat2026.0539.240.8
VEA-BrBackbone=GLM-4.1V-9B-Thinking2026.0639.2—
ADBackbone=Llama-2-13b-chat2026.053739.2
CGRBackbone=GLM-4.1V-9B-Thinking2026.0636.9—
VARBackbone=GLM-4.1V-9B-Thinking2026.0636.9—
DoLaBackbone=Llama-2-13b-chat2026.0536.839.6
DoLa+ADBackbone=Llama-2-13b-chat2026.0536.838.8
APCDBackbone=Llama-3-8b-it2026.0533.335.2
Llama-2-13b-chatDecoding Strategy=Greedy2026.0532.637.8
APCDBackbone=Llama-2-7b-chat2026.0529.428.4
DoLaBackbone=Llama-3-8b-it2026.0528.833.6
ADBackbone=Llama-3-8b-it2026.0528.833.4
DoLa+ADBackbone=Llama-3-8b-it2026.0528.633.3
Llama-3-8b-itDecoding Strategy=Greedy2026.0528.333.1
CADBackbone=Llama-3-8b-it2026.0526.530.1
DoLaBackbone=Llama-2-7b-chat2026.0525.427.4
ITIBackbone=Llama-2-13b-chat2026.0524.424.8
DoLa+ADBackbone=Llama-2-7b-chat2026.0524.326.3
ADBackbone=Llama-2-7b-chat2026.0524.226.1
CADBackbone=Llama-2-13b-chat2026.0523.726.6
Llama-2-7b-chatDecoding Strategy=Greedy2026.0523.425.5
ITIBackbone=Llama-2-7b-chat2026.0522.625.8
ITIBackbone=Llama-3-8b-it2026.0520.621.7
CADBackbone=Llama-2-7b-chat2026.0518.824.3