The paper investigates why Vision Language Models (VLMs) struggle with abstract reasoning tasks. Using the Relational Match-to-Sample paradigm and mechanistic analysis, the authors identify four factors—capability tier, model scale, number of objects per scene, and lack of per-object noise—that influence VLM performance. They uncover two competing neural circuits: an early surface-feature circuit and a late relational circuit, and show that ablating relational heads degrades performance more than random heads, suggesting the relational circuit is broadly used.
arXiv:2608.23074v1 Announce Type: new Abstract: Vision-language models (VLMs) can answer spatial questions, yet the mechanisms connecting object grounding to spatial reasoning remain poorly understoo...
arXiv:2607. 27703v2 Announce Type: replace Abstract: Vision-language models (VLMs) are increasingly used in embodied agents to interpret visual inputs, reason about spatial relationships, and make task-level decisions based on that reasoning.
arXiv:2605.13156v2 Announce Type: replace Abstract: Vision-language models (VLMs) have demonstrated remarkable capabilities in bridging visual perception and natural language understanding, enabling...
Chain-of-Thought (CoT) prompting is widely used as a test-time scaling strategy for Vision-Language Models (VLMs), but it remains unclear what is extended when VLMs generate longer reasoning traces. We ask whether CoT requires continued access to image tokens, or whether it mainly operates over visual information already made available earlier in the forward pass.
VBVR-Pro is a closed‑loop testbed that enables native visual reasoning through generation, offering 300 procedurally generated tasks that scale training and allow strong transfer to external benchmarks. It supplies verifiable reward scorers based on deterministic, task‑specific rules, outperforming VLM‑as‑a‑judge approaches and providing reliable signals for reinforcement learning. The suite also facilitates controlled modality studies, revealing that video generation excels at persistent spatiotemporal tracking while interleaved generation offers a compute‑efficient alternative, and highlights the importance of vision‑native trajectories for reasoning.
arXiv:2603. 23867v2 Announce Type: replace-cross Abstract: Vision-Language Models (VLMs) have been applied to a wide range of reasoning tasks, yet it remains unclear whether they can reason robustly under distribution shifts.
arXiv:2607. 12815v1 Announce Type: new Abstract: Chain-of-Thought (CoT) prompting is widely used as a test-time scaling strategy for Vision-Language Models (VLMs), but it remains unclear what is extended when VLMs generate longer reasoning traces.
arXiv:2508. 17298v3 Announce Type: replace-cross Abstract: Compositional visual reasoning has emerged as a key research frontier in multimodal AI, aiming to endow machines with the human-like ability to decompose visual scenes, ground intermediate concepts, and perform multi-step logical inference.
Vision-language models (VLMs) can answer spatial questions, yet the mechanisms connecting object grounding to spatial reasoning remain poorly understood. It is underexplored whether spatial reasoning...
arXiv:2609.26168v1 Announce Type: cross Abstract: Recent work reports that vision--language models (VLMs) struggle to establish and maintain stable reference in repeated reference games. Rather than...
arXiv:2606. 00148v1 Announce Type: cross Abstract: Multimodal large language models (MLLMs) often know the rule but pick the wrong answer: on abstract visual reasoning (AVR) tasks, a model can describe what it sees and name the underlying pattern, yet still fail to choose the matching candidate.
The paper introduces Percept-V, a dataset of 6,000 program-generated images across 30 domains designed to test simple visual perception skills from the TVPS-4 framework. Experiments show that state‑of‑the‑art multimodal large language models perform poorly compared to humans, especially as image complexity increases, and that fine‑tuning yields only limited generalization to related datasets. The study highlights specific perception skills that remain challenging for current models.