This work introduces Visualized Task Semantics (VTS), a controlled intervention that moves the question into the image while keeping the source problem and answer fixed, and requires no OCR or region metadata at inference.
Abstract
Multimodal large language models increasingly reason over screenshots and documents where the task itself may be written in pixels. Yet benchmarks usually place questions in text, leaving it unclear whether models use the same instruction equally well across channels. We introduce Visualized Task Semantics (VTS), a controlled intervention that moves the question into the image while keeping the source problem and answer fixed. Across six MLLMs and four benchmarks, accuracy drops in all 24 model-task pairs, by 17.8 points on average. Models often transcribe the visual question correctly yet fail to use it, exposing a semantic channel gap beyond OCR. To reduce this gap, we present prompt-region grounding, whose core design aligns the question region with typed semantics and recovers its clean representation from a masked view. At matched training cost, our method raises four-benchmark VTS accuracy from 58.0 to 66.3 while preserving accuracy on the original interface, and requires no OCR or region metadata at inference. Reading task-bearing text and grounding it as an instruction for reasoning are distinct capabilities.
Multimodal LLMs can see a document, but they often can't read it reliably. Small text, tables, visual cues, and topological elements still trip them up under direct visual inference, even when the page is already sitting in the model's context. Most document-VQA systems treat perception as fixed: they encode the page once, ask the question, and answer from whatever the model happened to extract in that single fast pass. We think document VQA needs slower, more deliberate perception: rather than answering from one fixed encoding, the model should spend a bit of extra compute at inference time working out what to look at next, and only then answer. We build this into \textbf{Q-Guide}, a small agent that reads a question, works out what evidence it is still missing, and calls targeted tool(s) to recover it---reading text where text is needed, zooming in where detail is needed, or grounding a region where position matters. On DocVQA2026 and Manga109, Q-Guide outperforms both direct prompting and recent multi-agent document systems ($65.0\%$ vs.\ $40.0\%$ on DocVQA2026, $32.4\%$ vs.\ $24.4\%$ on Manga109), and the improvement holds across three Claude backbones (Opus 4.6, Sonnet 4.6, and Opus 4.5). We find that accuracy scales with the perception budget---most of the gain appears within two to three deliberate rounds---and that the gain comes from directing perception to the right place, not from complex control logic: adding planners, routers, or multiple collaborating agents does not help.
Evaluating representative proprietary and open-source multimodal models, it is found that visual reasoning is strongly model- and environment-dependent, with no single setting consistently dominating across tasks.
Siyu Yan, Zhuoran Yan, Haiying Xu et al.· 0 citations
This work presents RegionSLM, a region-aware SLM designed to explicitly connect the question to its supporting regions, and curates ReDoc, a region-supervised corpus with 105k documents and 350k question-answer pairs obtained via a question-guided two-step filtering procedure.
Chao Wang, Hehe Fan, Huichen Yang et al.· Annual International ACM SIG...· 0 citations
ProVisE (Protocolized Visual Evaluation), a benchmark-agnostic framework that elicits protocol-constrained visual answers from image-generation models and parses them into structured predictions compatible with original metrics, is proposed and revealed, revealing complementary strengths of pixel-space expression and text-based reasoning.
Xu Wang, Kaixiang Yao, Miao Pan et al.· 1 citation
It is suggested that multimodal safety cannot be enforced reliably without modeling the requested operation, the visual target to which it applies, and the confidence of that grounding.
Recent advancements in Multimodal Large Language Models (MLLMs) have evolved from static perception to interleaved visual-language reasoning, often referred to as ``thinking with images''. A basic operation in this reasoning process is to zoom in on regions of interest (often represented with bounding boxes) to acquire finer visual details. In this paper, we propose \textbf{Seg}mentation before \textbf{Answer}ing (SegAnswer), which shifts the unit of zoom-in from the popular bounding box to pixel-level segmentation mask. By employing fine-grained masks to isolate the target area from cluttered environments, segmented visual input yields a more precise region of interest, effectively filtering out redundant background and interfering objects. Furthermore, the discrete patches of segmented visual input align more seamlessly with how MLLMs structure visual tokens via positional embeddings. In experiments, we evaluate SegAnswer across diverse benchmarks, including high-resolution perception, general perception, and hallucination. It achieves consistent improvements and also exhibits considerable performance on segmentation tasks, validating its capability for reliable pixel grounding.
Yake Wei, Yuan Wang, Fengyun Rao et al.· 0 citations