AI Paper Atlas

Research map / Benchmarks / ImageNet-Segmentation

ImageNet-Segmentation leaderboard

16 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on ImageNet-Segmentation. Best mIoU: 71.04 (ConceptAttention, ICML 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

ImageNet-Segmentation · mIoU

15 papers, each with its best reported number for its own method; higher is better.

5 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

ImageNet-Segmentation · mAP

11 papers, each with its best reported number for its own method; higher is better.

#mAP ↑MethodPaperTopic
194.29Ours
FLUX DiT
Attention (as Discrete-Time Markov) Chains
NeurIPS 2025 · Table: ImageNet segmentation
Attention and transformers
290.45ConceptAttention
Flux-Schnell
ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features
ICML 2025 · Table: ImageNet-Segmentation a
Fast diffusion sampling
386.4Mamba-Attr Ours
ViM-S
Explaining Modern Gated-Linear RNNs via a Unified Implicit Attention Formulation
ICLR 2025 · Table: Segmentation results
Attention and transformers
486.03Ours
ViT-base
Transformer Interpretability Beyond Attention Visualization
CVPR 2021 · Table 1
Attention and transformers
585.31SALF-CBM
ResNet-50
Show and Tell: Visually Explainable Deep Neural Nets via Spatially-Aware Concept Bottleneck Models
CVPR 2025 · Table 2
Recognition and long-tail
684.9Ours
CLIP ViT-B-32
Interpreting the Second-Order Effects of Neurons in CLIP
ICLR 2025 · Segmentation performance table
CLIP and zero-shot
782.1OpenCLIP ViT-B/14 with test-time register
OpenCLIP ViT-B/14
Vision Transformers Don't Need Trained Registers
NeurIPS 2025 · Zero-shot segmentation results
Attention and transformers
881.61Ours
CLIP
Interpreting CLIP's Image Representation via Text-Based Decomposition
ICLR 2024 · Table: Segmentation performanc
CLIP and zero-shot
980.18XBM-BLIP
BLIP
Explanation Bottleneck Models
AAAI 2025 · ImageNet Segmentation table
Human and embodied modeling
1079.9Ours
LLaVA
Interpreting and Editing Vision-Language Representations to Mitigate Hallucinations
ICLR 2025 · Segmentation Performance on Im
Multimodal LLMs
1 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.