Vendi Anomaly Score flags outliers via dataset diversity
The score needs no hyperparameter tuning, scales linearly, and led 148 embedding benchmarks plus ImageNet tests.
Stories per day · 30 days
4 RSS
Relateddiffusion4image-generation3pixel-space3representation-learning2ai-security1anomaly-detection1backdoor1clip1diffusion-models1dinov21embeddings1fid1flow-matching1flow-models1generative-models1generative-transformers1knowledge-distillation1model-reuse1
31 stories, newest first · one card per story
Stories tagged #imagenet
Patch-wise PCA links v-prediction failure to a bandwidth bottleneck
A Raw-injected Transformer restores effective v-prediction and is competitive on ImageNet 256×256.
SpecSlot raises pixel-space detail without longer global attention
On ImageNet-512, SpecSlot improves FID by 2.84 versus a FLOPs-matched JiT control and uses 3.9% fewer FLOPs.
Paper ties diffusion training speed to residual-gate layout
A teacher-free radial objective on ImageNet 256 with SiT-B/2 improves FID over dispersive loss while widening gate participation.
SpecSlot cuts global attention cost in pixel-space transformers
Only the lowest-frequency slot per spatial anchor attends globally; higher-frequency detail stays local.
TALOR rectifies open-set test-time features without backprop
On ImageNet-C with six csOOD datasets it beats UniEnt by 2.2% H-score while using 12% of the GPU memory.
Wiener-informed decoding speeds pixel-space flow model training
On ImageNet 256, PixelDiT-XL with WiD reports FID 1.53 after 500 epochs, matching prior quality with 37.5% fewer epochs.
Swift trains stronger diffusion models with a frozen weak denoiser
On one ImageNet-256 setup it beat the baseline's best FID with about half the training compute, including guide overhead.
Clean Data Can Still Carry Backdoors: Support-Persistent Backdoors for Model Reuse
DAR plants support-persistent backdoors that stay above 96 percent success after clean-data distillation.
Generative Representation Learning for Scalable Latent Tokenization
A generative tokenizer avoids representation collapse and reaches FID-50K 2.18 with JiT-L.
MixTok: Adaptive Visual Tokenization with Global 1D Tokens and Sparse 2D Refinement
MixTok mixes compact 1D tokens with sparse 2D refinement for budgeted visual tokenization.
Learning Representations to Guide Pixel Diffusion Transformers
Representation Guidance lets a 459M pixel diffusion model beat larger transformers, with FID 1.55.
One Block, Multiple Depths: Recurrent Vision Transformers with Depth-Programmed Experts
A recurrent vision Transformer matches full-depth accuracy with one block and about 70% fewer parameters.
Few-Step Generation via Data-Space Iteration
Data-space iteration enables few-step image generation without tuning a flow timestep schedule.
One Block, Multiple Depths: Recurrent Vision Transformers with Depth-Programmed Experts
A recurrent vision Transformer matches full-depth encoder accuracy with about 70% fewer stored parameters.
QuadTok: Quadtree Visual Tokenizer for Autoregressive Image Generation
QuadTok's quadtree tokenizer lets a 947M model reach 2.08 gFID while using fewer tokens.
Co-Evolving Paths and Flows via Path-Flow Alignment
Joint path-flow training with a stochastic regularizer improves ImageNet-256 FID without changing inference.
Data-free method distills diffusion models to one-step generation
Authors report ImageNet 256 FID of 2.04 at 1-NFE and 1.37 at 4-NFE after 40 epochs, without distillation data.
LiFT: Loop Flow Transformers
LiFT loops a shared Diffusion Transformer and lowers ImageNet FID while cutting parameters and FLOPs.
PaLoRA: Paced Low-Rank Adaptation for Continual Learning
PaLoRA paces LoRA continual-learning steps by effective rank, reducing forgetting on long task sequences.
Embedding Prediction Helps Image Generation
NEPA-DiT-XL predicts embeddings each denoising step and reaches ImageNet FID 1.32 with about one-third of REPA's compute.
DC-SAE: Deep Compression Semantic Autoencoder for Faster Diffusion Convergence
DC-SAE compresses images 32x and speeds diffusion training while beating DC-AE on ImageNet.
PSRben/VisionHOPE — new model trending #29 on Hugging Face
PSRben released VisionHOPE-T/S/B vision backbone weights for ImageNet, COCO, and ADE20K.
2
WISE-ATTA: When to Ask for Labels in Budgeted Active Test-Time Adaptation
WISE-ATTA schedules scarce labels during test-time adaptation and matches prior methods with fewer annotations.