{"ID":2843592,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.15717","arxiv_id":"2511.15717","title":"How Modality Shapes Perception and Reasoning: A Study of Error Propagation in ARC-AGI","abstract":"ARC-AGI and ARC-AGI-2 measure generalization-through-composition on small color-quantized grids, and their prize competitions make progress on these harder held-out tasks a meaningful proxy for systematic generalization. Recent instruction-first systems translate grids into concise natural-language or DSL rules executed in generate-execute-select loops, yet we lack a principled account of how encodings shape model perception and how to separate instruction errors from execution errors. We hypothesize that modality imposes perceptual bottlenecks -- text flattens 2D structure into 1D tokens while images preserve layout but can introduce patch-size aliasing -- thereby shaping which grid features are reliably perceived. To test this, we isolate perception from reasoning across nine text and image modalities using a weighted set-disagreement metric and a two-stage reasoning pipeline, finding that structured text yields precise coordinates on sparse features, images capture 2D shapes yet are resolution-sensitive, and combining them improves execution (about 8 perception points; about 0.20 median similarity). Overall, aligning representations with transformer inductive biases and enabling cross-validation between text and image yields more accurate instructions and more reliable execution without changing the underlying model.","short_abstract":"ARC-AGI and ARC-AGI-2 measure generalization-through-composition on small color-quantized grids, and their prize competitions make progress on these harder held-out tasks a meaningful proxy for systematic generalization. Recent instruction-first systems translate grids into concise natural-language or DSL rules execute...","url_abs":"https://arxiv.org/abs/2511.15717","url_pdf":"https://arxiv.org/pdf/2511.15717v1","authors":"[\"Bo Wen\",\"Chen Wang\",\"Erhan Bilal\"]","published":"2025-11-11T19:06:41Z","proceeding":"cs.AI","tasks":"[\"cs.AI\",\"cs.CV\",\"cs.MA\"]","methods":"[\"Transformer\"]","has_code":false}