Multimodal large language models can recognize visual content yet struggle to reason reliably over it across multiple steps. In counting, a model may see all objects but lose track of which ones it ha
Multimodal language models fail at multi-step visual reasoning due to a reference gap rather than perception, and task-matched external references like numbering significantly improve accuracy without training.
Multimodal large language models can recognize visual content yet struggle to reason reliably over it across multiple steps. In counting, a model may see all objects but lose track of which ones it has counted. We argue that such failures, commonly attributed to perception, reveal a reference gap: the failure to bind language to the relevant visual entities, regions, or locations. Motivated by the theory of visual routines, we test whether external references such as numbering can narrow this gap. We introduce ReGap-Bench, a paired benchmark of five visual reasoning tasks in which the scene, question, and answer are fixed and only the reference varies. Results show that task-matched…
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.
Diagnosing the Reference Gap: What Visual Scaffolds Can and Cannot Fix in Multimodal Reasoning was first seen on Hugging Face · Daily papers on Nov 3, 2026. Since then it has appeared in 13 checks on 2 different days. Its best position was #17 of 30, reached on Nov 3, 11:28 UTC. It dropped off the list after Nov 4, 03:28 UTC.
Position on Hugging Face · Daily papers at every check (lower is better).
Best rank per day (UTC).
Platform metric as reported by Hugging Face.
Archived daily, weekly and monthly lists that include this entry.
Not at the moment. It was last on the list on Nov 4, 03:28 UTC, after 2 days of trending.
#17 on Hugging Face · Daily papers, on Nov 3, 11:28 UTC. Trendoday checks this list every 120 minutes.
It was first seen on Nov 3, 2026 and has been on the list on 2 days (13 checks) since.