Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence
ReaLVR adds visual-evidence supervision to latent reasoning, hitting 63.7% on Qwen2.5-VL-7B and scaling to 235B.
1 stories, newest first · one card per story
ReaLVR adds visual-evidence supervision to latent reasoning, hitting 63.7% on Qwen2.5-VL-7B and scaling to 235B.
Today 795 readers · 1K views · 795 reads · all time 159K views