paper-with-me

Papers

Edit Spillover as a Probe: Do Image Editing Models Implicitly Understand World Relations?

2026-03-18 · Guandong Li, Zhaobin Chu arxiv

Instruction-following image editing models are expected to modify only the specified region while keeping the rest of the image unchanged. However, in practice, we observe a pervasive phenomenon -- edit spillover: models alter semantically related but unspecified content outside the edit region. This raises a fundamental question -- does spillover reflect genuine implicit world understanding, or is it merely attention leakage? We propose EditSpilloverProbe, a systematic framework that repurposes edit spillover as a natural probe for world knowledge in image editing models. We introduce a spillover taxonomy (spatial, semantic, mixed, random), an automated detection-and-classification pipeline, and a benchmark dataset constructed from real-world Chinese text editing tasks, EditSpilloverBench. Systematic evaluation of 5 representative editing models reveals three core findings: (1) spillover rates vary dramatically across architectures, from 3.49% to 11.46%, with a 3.3x ratio; (2) absolute semantic spillover quantity reveals models' world understanding capability -- nano_banana produces the most semantic spillover (27.8 per image), while qwen_2511 has the most precise editing control but lower semantic spillover (16.3 per image), revealing a trade-off between editing control and world understanding; (3) spatial decay analysis shows spillover area density decays exponentially with distance, but the proportion of semantically relevant spillover remains constant (40%-58%), providing direct evidence that semantic spillover reflects genuine world understanding rather than spatial diffusion.

📄 PDF Abstract BibTeX arXiv:2603.17876

Code (0)

등록된 구현이 없습니다.

Tasks

Image Editing

Similar Papers 제목 키워드 기반

Edit Fidelity Field: Semantics-Aware Region Isolation for Training-Free Scene Text Editing

2026-04-19 · Guandong Li, Mengxia Ye arxiv

Scene text editing (STE) has achieved remarkable progress in accurately rendering target text through diffusion-based methods. However, we identify a critical yet overlooked problem: edit spillover -- when editing a targ…

Evaluating Image Editing with LLMs: A Comprehensive Benchmark and Intermediate-Layer Probing Approach

2026-03-20 · Shiqi Gao, Zitong Xu, Kang Fu, Huiyu Duan 외 arxiv

Evaluating text-guided image editing (TIE) methods remains a challenging problem, as reliable assessment should simultaneously consider perceptual quality, alignment with textual instructions, and preservation of origina…

Image Editing

Toward 360-Degree Indoor Panorama Editing via Tuning-Free Diffusion Model with Refocusing Cross-Attention

2026-06-12 · Dinh-Khoi Vo, Nhut-Thanh Le-Hinh, Viet-Tham Huynh, Tam V. Nguyen 외 arxiv

Zero-shot text-guided diffusion has significantly advanced image editing; however, its practical usability remains constrained by three persistent challenges: prompt brittleness that requires meticulous prompt engineerin…

Image ManipulationPrompt EngineeringImage Editing

Joint Knowledge Editing for Information Enrichment and Probability Promotion

2024-12-22 · Wenhang Shi, Yiren Chen, Shuqing Bian, Xinyi Zhang 외

Knowledge stored in large language models requires timely updates to reflect the dynamic nature of real-world information. To update the knowledge, most knowledge editing methods focus on the low layers, since recent pro…

counterfactualknowledge editing

Do Image Editing Models Understand Lighting?

2026-06-25 · Tim Küchler, Johann-Friedrich Feiden, Matthias Nießner, Carsten Rother arxiv

While recent advancements in generative image editing models have achieved stunning visual fidelity, it remains an open question whether these systems possess an intrinsic knowledge of real-world lighting. Existing bench…

Image Editing