Where to Refine, When to Stop: Rethinking Redundancy via Latent Discrepancy for Efficient Visual Autoregressive Generation

• URL: http://arxiv.org/abs/2606.00310v1
• Date: Fri, 29 May 2026 19:34:39 GMT
• Title: Where to Refine, When to Stop: Rethinking Redundancy via Latent Discrepancy for Efficient Visual Autoregressive Generation
• Abstract summary: We propose a training-free framework that removes redundancy via latent discrepancy.<n> LD-Pruning substantially reduces inference while maintaining high generation quality, achieving up to 2.35x speedup on Infinity-8B.
• Score: 50.139984798361375
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Visual Autoregressive (VAR) models deliver high-quality image generation but suffer from significant inference latency at high resolutions. Recent acceleration approaches most rely on heuristic measures with layer features to prune tokens. Such heuristics are sensitive to complex contextual semantics, leading to inaccurate identification of redundant computation and poor adaptability across prompts. We rethink redundancy in VAR from the perspective of its impact on pixel-space generation and introduce Latent Discrepancy. This unified metric quantifies a token's contribution by measuring the change in model states during generation. Our analysis shows that redundancy is more accurately identified when guided by image latent or pixel-space signals. We further observed that in classifier-free guidance (CFG), the convergence trend of the discrepancy between conditional and unconditional branches exhibits high dynamics with different prompts. Based on these findings, we propose LD-Pruning (Latent Discrepancy Pruning), a training-free framework that removes redundancy via latent discrepancy by integrating decoding-free region selection and adaptive unconditional-branch skipping. Extensive experiments show that LD-Pruning substantially reduces inference latency while maintaining high generation quality, achieving up to 2.35x speedup on Infinity-8B.

Related papers

• Balancing Emotional Alignment and Semantic Consistency in Image Generation via Reinforcement Learning with Valence-Arousal Anchoring [55.57242007016976]
  Continuous emotion control in text-to-image generation requires a model to improve affective alignment without changing the objects, layout, or scene described by the prompt.<n>Existing supervised emotion-injection methods often optimize feature-space proxies and may therefore exhibit emotion-semantic drift.<n>We address this problem with a flow-matching image-generation framework that combines continuous valence-arousal conditioning, Group Relative Policy Optimization, and a neutral semantic anchor.
  arXiv  Detail & Related papers  (2026-09-11T13:24:35Z)
• Local Epistemic Uncertainty Guided Active Sampling for Plug-and-play Diffusive Image Restoration [15.371827475537776]
  Local Epistemic Uncertainty Guided Active Sampling framework is developed.<n>It balances detail preservation and artifact suppression in the spatial domain.<n>It reduces sampling time with negligible memory overhead.
  arXiv  Detail & Related papers  (2026-08-07T09:00:20Z)
• Where Does Generative Difficulty Reside? An Empirical Study of Target Representations [62.54876287800644]
  The target representation defines the distribution an image generator must learn, yet it is often treated as an interchangeable interface.<n>We study raw pixels, SD-VAE latents and DINOv2 as well as MAE representation-autoencoder features within a unified masked autoregressive rectified-flow model.<n>Our results show that compression, reconstruction fidelity, token dimensionality, and visible semantic clustering do not individually predict generative behavior.
  arXiv  Detail & Related papers  (2026-08-01T12:32:31Z)
• Rarity-Aware Discrete Diffusion with Spatially Consistent Decoding for Photo-Realistic Image Super-Resolution [80.28905986756685]
  DiMOO-SR is a rarity-aware multimodal discrete diffusion framework for photo-realistic image Super-Resolution.<n>Experiments on widely used real-world SR benchmarks demonstrate that DiMOO-SR achieves competitive perceptual quality with only a few parallel decoding steps.
  arXiv  Detail & Related papers  (2026-07-20T07:01:01Z)
• Unlocking Complex Visual Generation via Closed-Loop Verified Reasoning [11.461705552241163]
  Closed-Loop Visual Reasoning (CLVR) is a comprehensive system that couples visual-language logical planning with pixel-level diffusion generation.<n> CLVR introduces an automated data engine with step-level visual verification to synthesize reliable reasoning trajectories.<n>To mitigate the severe latency bottleneck caused by iterative denoising, we propose $$-Space Weight (DSWM)
  arXiv  Detail & Related papers  (2026-05-14T14:22:25Z)
• CASCADE: Context-Aware Relaxation for Speculative Image Decoding [9.004650208708235]
  Autoregressive generation is a powerful approach for high-fidelity image synthesis.<n>Current approaches fail to achieve efficiency gains comparable to those observed in text generation.<n>We identify previously overlooked patterns in the target model's behavior that emerge naturally in tree-based speculative decoding.
  arXiv  Detail & Related papers  (2026-05-08T04:32:17Z)
• CIAR: Interval-based Collaborative Decoding for Image Generation Acceleration [47.52310736101766]
  Auto-regressive (AR) models have recently made notable progress in image generation, achieving performance comparable to diffusion-based approaches.<n>We address this via a cloud-device collaboration framework textbfCIAR, which utilizes on-device self-verification to handle two key properties of visual synthesis.<n>Our solution centers on an on-device token uncertainty quantifier, which adopts continuous probability intervals to accelerate processing and make it feasible for large visual vocabularies.
  arXiv  Detail & Related papers  (2026-03-26T14:04:50Z)
• LADR: Locality-Aware Dynamic Rescue for Efficient Text-to-Image Generation with Diffusion Large Language Models [10.535207607235046]
  Locality-Aware Dynamic Rescue (LADR) is a training-free method that expedites inference by exploiting the spatial Markov property of images.<n>LADR prioritizes the recovery of tokens at the ''generation frontier'', regions spatially adjacent to observed pixels.<n>Experiments on four text-to-image generation benchmarks demonstrate that our LADR achieves an approximate 4 x speedup over standard baselines.
  arXiv  Detail & Related papers  (2026-03-13T15:12:41Z)
• Guiding Diffusion-based Reconstruction with Contrastive Signals for Balanced Visual Representation [81.40978077888693]
  Contrastive Language-Image Pre-training (CLIP) has become a key bottleneck for downstream performance.<n>Recent solutions use diffusion models to enhance representations by conditioning image reconstruction on CLIP visual tokens.<n>We integrate contrastive signals into diffusion-based reconstruction to pursue more comprehensive visual representations.
  arXiv  Detail & Related papers  (2026-03-05T04:45:49Z)
• Multi-Scale Local Speculative Decoding for Image Generation [10.239314110594249]
  We introduce Multi-Scale Local Speculative Decoding (MuLo-SD)<n>MuLo-SD combines multi-resolution drafting with spatially informed verification to accelerate AR image generation.<n>We demonstrate that MuLo-SD achieves substantial speedups up to $mathbf1.7times$.
  arXiv  Detail & Related papers  (2026-01-08T17:39:35Z)
• Training-Free Generation of Diverse and High-Fidelity Images via Prompt Semantic Space Optimization [50.5332987313297]
  We propose Token-Prompt embedding Space Optimization (TPSO), a training-free and model-agnostic module.<n>TPSO introduces learnable parameters to explore underrepresented regions of the token embedding space, reducing the tendency of the model to repeatedly generate samples from strong modes of the learned distribution.<n>In experiments on MS-COCO and three diffusion backbones, TPSO significantly enhances generative diversity, improving baseline performance from 1.10 to 4.18 points, without sacrificing image quality.
  arXiv  Detail & Related papers  (2025-11-25T00:42:09Z)
• Rethinking Discrete Tokens: Treating Them as Conditions for Continuous Autoregressive Image Synthesis [79.98107530577576]
  DisCon is a novel framework that reinterprets discrete tokens as conditional signals rather than generation targets.<n>DisCon achieves a gFID score of 1.38 on ImageNet 256$times $256 generation, outperforming state-of-the-art autoregressive approaches by a clear margin.
  arXiv  Detail & Related papers  (2025-07-02T14:33:52Z)
• Multi-Modality Driven LoRA for Adverse Condition Depth Estimation [61.525312117638116]
  We propose Multi-Modality Driven LoRA (MMD-LoRA) for Adverse Condition Depth Estimation.<n>It consists of two core components: Prompt Driven Domain Alignment (PDDA) and Visual-Text Consistent Contrastive Learning (VTCCL)<n>It achieves state-of-the-art performance on the nuScenes and Oxford RobotCar datasets.
  arXiv  Detail & Related papers  (2024-12-28T14:23:58Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.