arXiv AI
CoTA++: Understanding the Repeat Curse in dMLLMs from an Information Flow Perspective
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv AI
Jun 3
dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Caching
arXiv:2506. 06295v2 Announce Type: replace-cross Abstract: Autoregressive Models (ARMs) have long dominated the landscape of Large Language Models.
arXiv AI
Sep 1
Tail-Replay: Escaping the Curse of Linear Attention in Prefix Caching for Hybrid LLMs
arXiv:2608.30310v1 Announce Type: cross Abstract: Hybrid large language models interleave full-attention layers with linear-attention layers to reduce the cost of long-context inference. This structu...
arXiv Machine Learning
Aug 7
QEvict: Recoverable Quantized KV Eviction for Attention-Drift-Robust Long-Context Decoding
arXiv:2608. 05326v1 Announce Type: new Abstract: Autoregressive large language model inference is increasingly constrained by the memory footprint of the Key-Value (KV) cache.
arXiv Machine Learning
Sep 16
Window-Diffusion: Accelerating Diffusion Language Model Inference with Windowed Token Pruning and Caching
The paper introduces Window-Diffusion, a method that accelerates diffusion language model inference by pruning and caching tokens within a sliding window. It categorizes undecoded tokens into active, buffer, and far-field groups, computing only the first two while discarding the rest. Experiments on LLaDA and Dream demonstrate up to 99× speedup with minimal loss in generation quality.
arXiv Computation and Language
Aug 21
Remask, Don't Replace: Token-to-Mask Refinement in Diffusion Language Models
arXiv:2604. 18738v3 Announce Type: replace Abstract: Diffusion language models (dLLMs) generate text through iterative denoising, filling multiple masked positions at each step.
arXiv Machine Learning
Aug 28
Affix Cache for Diffusion Large Language Models
The paper introduces ACache, an affix-oriented cache reuse mechanism for Diffusion Large Language Models (DLLMs). ACache identifies a small set of critical affix tokens, called Anchor Tokens, and selectively recomputes their key-value states while reusing the rest of the affix cache. Experiments on Fast-dLLM and Nano-vLLM show that recomputing about 20% of affix tokens restores accuracy and can reduce recompute latency by up to 55.7% while improving throughput by up to 1.68×.