# Mind the Gap: Diagnosing Constraint Discovery Failures in Text-in-Image Editing

arXiv: https://arxiv.org/abs/2606.15982 · HTML: https://mlsanity.com/paper/2606.15982/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 17 of its 92 references that are in the map.
- First posted on arXiv 14 Jun 2026.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)

## Builds on (most cited first, 17 of 17)

- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [Discovering Latent Knowledge in Language Models Without Supervision](https://mlsanity.com/paper/2212.03827/) (2022)
- [Multimodal Chain-of-Thought Reasoning in Language Models](https://mlsanity.com/paper/2302.00923/) (2023)
- [Decomposed Prompting: A Modular Approach for Solving Complex Tasks](https://mlsanity.com/paper/2210.02406/) (2022)
- [Set-of-Mark Prompting Unleashes Extraordinary Visual Grounding in GPT-4V](https://mlsanity.com/paper/2310.11441/) (2023)
- [On the Binding Problem in Artificial Neural Networks](https://mlsanity.com/paper/2012.05208/) (2020)
- [EditVal: Benchmarking Diffusion Based Text-Guided Image Editing Methods](https://mlsanity.com/paper/2310.02426/) (2023)
- [ReasonEdit: Towards Reasoning-Enhanced Image Editing Models](https://mlsanity.com/paper/2511.22625/) (2025)
- [TextEditBench: Evaluating Reasoning-aware Text Editing Beyond Rendering](https://mlsanity.com/paper/2512.16270/) (2025)
- [GRADE: Benchmarking Discipline-Informed Reasoning in Image Editing](https://mlsanity.com/paper/2603.12264/) (2026)
- [CRAFT: Continuous Reasoning and Agentic Feedback Tuning for Multimodal Text-to-Image Generation](https://mlsanity.com/paper/2512.20362/) (2025)
- [WorldEdit: Towards Open-World Image Editing with a Knowledge-Informed Benchmark](https://mlsanity.com/paper/2602.07095/) (2026)
- [Is This Edit Correct? A Multi-Dimensional Benchmark for Reasoning-Aware Image Editing](https://mlsanity.com/paper/2606.05172/) (2026)
- [VDE Bench: Evaluating The Capability of Image Editing Models to Modify Visual Documents](https://mlsanity.com/paper/2602.00122/) (2026)
- [Beyond Perception Errors: Semantic Fixation in Large Vision-Language Models](https://mlsanity.com/paper/2604.12119/) (2026)

## Abstract (arXiv)

A key challenge in multimodal reasoning is determining which visual dependencies become relevant under a specific task, rather than merely recognizing visible content. We study this through edit-induced constraint discovery in text-in-image editing, a controlled diagnostic setting where a local text change can activate secondary consistency constraints: given a valid editing instruction and an image, can a model identify the secondary regions that must also change? Across 461 diagnostic cases, four MLLMs, and 19 constraint subtypes, models recover only 46% case-level macro recall under unguided prompting versus 94% when constraints are explicitly provided, suggesting that a substantial portion of the failure arises when models must decide which unstated dependencies to surface. Oracle-field decomposition shows that case-specific causal explanations are the most effective partial guidance (0.782 recall), above region names (0.610) or type labels (0.646), suggesting that edit-specific causal cues account for much of the oracle gain. A downstream experiment further shows that higher self-discovery recall does not necessarily improve task performance: unverified self-discovery introduces false positives that offset recall gains, motivating precision-aware constraint elicitation.