Thinking with Drafting: Optical Decompression via Logical Reconstruction

ACL ARR 2026 January Submission3594 Authors

Keywords: Visual Algebra Reasoning, Optical Decompression, Graphic Intermediate Representation, Multimodal Reasoning, Neuro-Symbolic Alignment

Abstract: Existing multimodal large language models have achieved high-fidelity visual perception and exploratory visual generation. However, a precision paradox persists in complex reasoning tasks: optical perception systems transcribe symbols without capturing logical topology, while pixel-based generative models produce visual artifacts lacking mathematical exactness. To bridge this gap, we propose that reasoning over visual inputs be reconceptualized as \textit{optical decompression}—the process of reconstructing latent logical structures from compressed visual tokens. Guided by the axiom that \textit{Parsing is Reasoning}, we introduce Thinking with Drafting (TwD), which utilizes a minimalist Domain-Specific Language (DSL) as a grounding intermediate representation. Instead of directly generating answers, TwD forces the model to draft its mental model into DSL, which can render geometric verification. Using the bar model as a testbed, our experiments demonstrate that structured DSLs serve as a superior cognitive scaffold. Our work establishes a closed-loop system where visual generation acts not as a creative output but as a deterministic verifier, offering a generalizable path for grounding abstract reasoning.

Paper Type: Long

Research Area: Multimodality and Language Grounding to Vision, Robotics and Beyond

Research Area Keywords: multimodality, cross-modal application, vision question answering

Contribution Types: NLP engineering experiment, Data resources, Data analysis

Languages Studied: English

Submission Number: 3594