mlsanityData to 13 Oct 2026

WaveFilter: Enhancing the Long-Context Capability of Diffusion LLMs via Wavelet-Guided KV Cache Filtering

arXiv 2606.00724 ↗First posted 30 May 2026cs.CL cs.AIcited by 1

On the map: Efficient LLM inference and serving → Diffusion LLM and MoE inference

Abstract (arXiv)

Diffusion Large Language Models (DLMs) have demonstrated significant advantages across various tasks. However, constrained by their multi-step iterative inference mechanism, their computational overhead and inference latency in long-context tasks have become core bottlenecks restricting their large-scale deployment. When processing long sequences, existing Key-Value (KV) caching mechanisms often face a dilemma where generation quality degrades drastically, where the core challenge lies in precisely and efficiently filtering critical tokens within ultra-long contexts. Inspired by the human reading process, we propose WaveFilter, a universal and training-free caching framework. This framework innovatively introduces the wavelet transform for decomposition of long sequences to achieve precise identification of key tokens, based on which a sparse KV Cache is constructed to compute the final contextual representation. Experimental results demonstrate that WaveFilter, as a plug-and-play generic framework, significantly enhances the performance of existing mainstream KV Cache methods in complex long-context tasks.

Key facts. Cited by 1 paper in the map (data through 13 Oct 2026). Placed in Diffusion LLM and MoE inference, one of 12 directions in Efficient LLM inference and serving. Builds on 26 of its 34 references that are in the map. First posted on arXiv 30 May 2026.

Builds on

26 of 34 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Attention Is All You Need for KV Cache in Diffusion LLMs16 Oct
  2. 2025d2Cache: Accelerating Diffusion-Based LLMs via Dual Adaptive Caching27 Sep
  3. 2025Sparse-dLLM: Accelerating Diffusion LLMs with Dynamic Cache Eviction4 Aug
  4. 2025Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding28 May

Other references · 22

  1. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,213
  2. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,023
  3. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,077
  4. 2023Efficient Streaming Language Models with Attention Sinkscited by 2,137
  5. 2021Structured Denoising Diffusion Models in Discrete State-Spacescited by 1,685
  6. 2023LongBench: A Bilingual, Multitask Benchmark for Long Context Understandingcited by 1,386
  7. 2024RULER: What's the Real Context Size of Your Long-Context Language Models?cited by 1,302
  8. 2022Diffusion-LM Improves Controllable Text Generationcited by 1,012
  9. 2024SnapKV: LLM Knows What You are Looking for Before Generationcited by 856
  10. 2022DiffuSeq: Sequence to Sequence Text Generation with Diffusion Modelscited by 444
  11. 2022A Continuous Time Framework for Discrete Denoising Modelscited by 442
  12. 2023DeepCache: Accelerating Diffusion Models for Freecited by 353
10 more
  1. 2023Photorealistic Video Generation with Diffusion Modelscited by 305
  2. 2024DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Headscited by 284
  3. 2025DiffuCoder: Understanding and Improving Masked Diffusion Models for Code Generationcited by 184
  4. 2025dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Cachingcited by 168
  5. 2025Fast-dLLM v2: Efficient Block-Diffusion LLMcited by 132
  6. 2022Wave-ViT: Unifying Wavelet and Transformers for Visual Representation Learningcited by 78
  7. 2023Diffusion Models With Learned Adaptive Noisecited by 51
  8. 2025LongLLaDA: Unlocking Long Context Capabilities in Diffusion LLMscited by 38
  9. 2025Tensor Product Attention Is All You Needcited by 28
  10. 2025Learnable Multi-Scale Wavelet Transformer: A Novel Alternative to Self-Attentioncited by 7

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Affix Cache for Diffusion Large Language Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025dKV-Cache: The Cache for Diffusion Language Models
  2. 2026GroupKV: Hierarchical KV Cache Management for Long-Context Diffusion LLM Inference
  3. 2024KV Cache Compression, But What Must We Give in Return? A Comprehensive Benchmark of Long Context Capable Approaches
  4. 2026EntropyCache: Decoded Token Entropy Guided KV Caching for Diffusion Language Models
  5. 2026Polestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs
  6. 2025FlashDLM: Accelerating Diffusion Language Model Inference via Efficient KV Caching and Guided Diffusion
  7. 2026Window-Diffusion: Accelerating Diffusion Language Model Inference with Windowed Token Pruning and Caching
  8. 2026Mosaic: Unlocking Long-Context Inference for Diffusion LLMs via Global Memory Planning and Dynamic Peak Taming