List your company

Visual Tool-Use on HRBench 8K

73.7AccuracyARM-Thinker-7B
57.68461.8426670.158Nov 24, 2025Nov 25, 2025Nov 27, 2025Nov 29, 2025Nov 30, 2025Dec 2, 2025Dec 4, 2025

Evaluation results

MethodLinks
ARM-Thinker-7BSize=7B, Backbone=Qwen2.5-VL-7B2025.1273.7—
Mini-o3Source=Lai et al. (2025)2025.1273.3—
CodeV-7B-RLBackbone=7B, Training=RL2025.1171.313.3
Thyme-RL-7BBackbone=7B, Training=RL2025.1171.21.2
Qwen3-VL-8BSize=8B2025.1270.4—
InternVL3.5-8BSize=8B2025.1269.9—
DeepEyesSource=Lai et al. (2025)2025.1269.5—
DeepEyes-7BBackbone=7B2025.1169.16.7
Pixel-Reasoner-7BBackbone=7B2025.1168.57.6
InternVL3-8BSize=8B2025.1268.4—
Pixel ReasonerSource=Lai et al. (2025)2025.1266.9—
Qwen2.5-VL-7BSize=7B2025.1264.6—
GPT-4oSource=Lai et al. (2025)2025.1258.3—