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

arXiv: https://arxiv.org/abs/2603.19775 · HTML: https://mlsanity.com/paper/2603.19775/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 52 of its 82 references that are in the map.
- First posted on arXiv 20 Mar 2026.

## Builds on (most cited first, 30 of 52)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://mlsanity.com/paper/2201.12086/) (2022)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Scaling Rectified Flow Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2403.03206/) (2024)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [InstructPix2Pix: Learning to Follow Image Editing Instructions](https://mlsanity.com/paper/2211.09800/) (2022)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [CLIPScore: A Reference-free Evaluation Metric for Image Captioning](https://mlsanity.com/paper/2104.08718/) (2021)
- [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025)
- [Transformer Feed-Forward Layers Are Key-Value Memories](https://mlsanity.com/paper/2012.14913/) (2020)
- [Imagic: Text-Based Real Image Editing with Diffusion Models](https://mlsanity.com/paper/2210.09276/) (2022)
- [ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2304.05977/) (2023)
- [Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generation](https://mlsanity.com/paper/2305.01569/) (2023)
- [MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing](https://mlsanity.com/paper/2304.08465/) (2023)
- [MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editing](https://mlsanity.com/paper/2306.10012/) (2023)
- [MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessment](https://mlsanity.com/paper/2204.08958/) (2022)
- [NIMA: Neural Image Assessment](https://mlsanity.com/paper/1709.05424/) (2017)
- [AdaLoRA: Adaptive Budget Allocation for Parameter-Efficient Fine-Tuning](https://mlsanity.com/paper/2303.10512/) (2023)
- [Emu Edit: Precise Image Editing via Recognition and Generation Tasks](https://mlsanity.com/paper/2311.10089/) (2023)
- [Text2LIVE: Text-Driven Layered Image and Video Editing](https://mlsanity.com/paper/2204.02491/) (2022)
- [Blind Image Quality Assessment Using A Deep Bilinear Convolutional Neural Network](https://mlsanity.com/paper/1907.02665/) (2019)
- [Blind Image Quality Assessment via Vision-Language Correspondence: A Multitask Learning Perspective](https://mlsanity.com/paper/2303.14968/) (2023)
- [EDICT: Exact Diffusion Inversion via Coupled Transformations](https://mlsanity.com/paper/2211.12446/) (2022)
- [An Edit Friendly DDPM Noise Space: Inversion and Manipulations](https://mlsanity.com/paper/2304.06140/) (2023)

## Built on by (most cited first, 3 of 3)

- [What Can I Edit? Open-Ended Strategy Discovery and the Emotion Editability Landscape](https://mlsanity.com/paper/2607.23920/) (2026)
- [Edit2TikZ: A Comprehensive and Challenging Benchmark for Scientific Figure Editing with TikZ](https://mlsanity.com/paper/2608.13441/) (2026)
- [Do MLLM Judges Judge the Edit? Auditing Bias in Image Editing Evaluation with Verified Quality Preservation](https://mlsanity.com/paper/2610.01670/) (2026)

## Abstract (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 original image content. Despite rapid progress in TIE models, existing evaluation benchmarks remain limited in scale and often show weak correlation with human perceptual judgments. In this work, we introduce TIEdit, a benchmark for systematic evaluation of text-guided image editing methods. TIEdit consists of 512 source images paired with editing prompts across eight representative editing tasks, producing 5,120 edited images generated by ten state-of-the-art TIE models. To obtain reliable subjective ratings, 20 experts are recruited to produce 307,200 raw subjective ratings, which accumulates into 15,360 mean opinion scores (MOSs) across three evaluation dimensions: perceptual quality, editing alignment, and content preservation. Beyond the benchmark itself, we further propose EditProbe, an LLM-based evaluator that estimates editing quality via intermediate-layer probing of hidden representations. Instead of relying solely on final model outputs, EditProbe extracts informative representations from intermediate layers of multimodal large language models to better capture semantic and perceptual relationships between source images, editing instructions, and edited results. Experimental results demonstrate that widely used automatic evaluation metrics show limited correlation with human judgments on editing tasks, while EditProbe achieves substantially stronger alignment with human perception. Together, TIEdit and EditProbe provide a foundation for more reliable and perceptually aligned evaluation of text-guided image editing methods.