DeVAR is the first application of visual autoregressive modeling to low-dose CT denoising, using next-scale token prediction, a residual refiner, and hybrid discrete-continuous decoding to outperform prior methods on two public datasets.
Hart: Efficient visual generation with hybrid autoregressive transformer
10 Pith papers cite this work. Polarity classification is still indexing.
10
Pith papers citing it
citation-role summary
background 1
baseline 1
citation-polarity summary
verdicts
UNVERDICTED 10representative citing papers
HACK++ is a head-aware KV cache compression framework for VAR models that decouples current-scale attention from historical cache under adaptive per-head budgets to achieve near-lossless generation at 30% attention and 10% cache budgets.
LD-Pruning applies latent discrepancy to prune tokens and adaptively skip unconditional branches in VAR models for up to 2.35x faster inference with preserved quality.
ExtraVAR enables resolution extrapolation in visual autoregressive models by stage-aware RoPE remapping and entropy-driven attention scaling, suppressing repetition and detail loss.
Prologue adds a small set of learnable tokens trained exclusively with AR cross-entropy loss to decouple generation from reconstruction in autoregressive visual models, yielding lower gFID on ImageNet 256x256.
Discrete Stochastic Localization lets a single trained network support an entire family of per-token SNR paths for discrete sequence generation, with masked diffusion as a special case, and improves MAUVE scores when fine-tuning pretrained checkpoints.
PacTure uses view packing and next-scale autoregressive prediction to generate consistent multi-view PBR textures faster than prior sequential or cross-attention methods.
A transcoder-based in-place replacement of the bottleneck layer enables selective concept removal in modern diffusion and autoregressive image models without degrading output quality.
VarKD is a distillation framework for visual AR models that uses student samples and selective teacher supervision to reduce token ambiguity, outperforming prior baselines on ImageNet.
FlashAR accelerates autoregressive image generation up to 22.9x by post-training a pre-trained raster-scan model with a complementary vertical head and dynamic fusion for two-way next-token prediction.
citing papers explorer
-
DeVAR: Low-Dose CT Denoising via Visual Autoregressive Modeling
DeVAR is the first application of visual autoregressive modeling to low-dose CT denoising, using next-scale token prediction, a residual refiner, and hybrid discrete-continuous decoding to outperform prior methods on two public datasets.
-
HACK++: Towards More Effective Head-Aware Key-Value Compression for Efficient Visual Autoregressive Modeling
HACK++ is a head-aware KV cache compression framework for VAR models that decouples current-scale attention from historical cache under adaptive per-head budgets to achieve near-lossless generation at 30% attention and 10% cache budgets.
-
Where to Refine, When to Stop: Rethinking Redundancy via Latent Discrepancy for Efficient Visual Autoregressive Generation
LD-Pruning applies latent discrepancy to prune tokens and adaptively skip unconditional branches in VAR models for up to 2.35x faster inference with preserved quality.
-
ExtraVAR: Stage-Aware RoPE Remapping for Resolution Extrapolation in Visual Autoregressive Models
ExtraVAR enables resolution extrapolation in visual autoregressive models by stage-aware RoPE remapping and entropy-driven attention scaling, suppressing repetition and detail loss.
-
Autoregressive Visual Generation Needs a Prologue
Prologue adds a small set of learnable tokens trained exclusively with AR cross-entropy loss to decouple generation from reconstruction in autoregressive visual models, yielding lower gFID on ImageNet 256x256.
-
Discrete Stochastic Localization for Non-autoregressive Generation
Discrete Stochastic Localization lets a single trained network support an entire family of per-token SNR paths for discrete sequence generation, with masked diffusion as a special case, and improves MAUVE scores when fine-tuning pretrained checkpoints.
-
PacTure: Efficient PBR Texture Generation on Packed Views with Visual Autoregressive Models
PacTure uses view packing and next-scale autoregressive prediction to generate consistent multi-view PBR textures faster than prior sequential or cross-attention methods.
-
Concept Removal for Frontier Image Generative Models
A transcoder-based in-place replacement of the bottleneck layer enables selective concept removal in modern diffusion and autoregressive image models without degrading output quality.
-
Knowledge Distillation for Visual Autoregressive Models
VarKD is a distillation framework for visual AR models that uses student samples and selective teacher supervision to reduce token ambiguity, outperforming prior baselines on ImageNet.
-
FlashAR: Efficient Post-Training Acceleration for Autoregressive Image Generation
FlashAR accelerates autoregressive image generation up to 22.9x by post-training a pre-trained raster-scan model with a complementary vertical head and dynamic fusion for two-way next-token prediction.