Universal Assisted Generation: Faster Decoding with Any Assistant Model

How to generate text: using different decoding methods for language generation with Transformers

Assisted Generation: a new direction toward low-latency text generation

DC-Leap: Training-Free Acceleration of dLLMs via Draft-Guided Contiguous Leaping Decoding

arXiv:2607. 20467v1 Announce Type: new Abstract: While parallel decoding is central to the efficiency of Diffusion Large Language Models (dLLMs), current strategies are often hindered by overly conservative confidence thresholds.

Faster Text Generation with Self-Speculative Decoding

Gumbel Distillation for Parallel Text Generation

arXiv:2603. 22216v2 Announce Type: replace-cross Abstract: The slow, sequential nature of autoregressive (AR) language models has driven the adoption of parallel decoding methods.

WAND: Windowed Attention and Knowledge Distillation for Efficient Autoregressive Text-to-Speech Models

arXiv:2604. 08558v2 Announce Type: replace-cross Abstract: Recent decoder-only autoregressive text-to-speech (AR-TTS) models produce high-fidelity speech, but their memory and compute costs scale quadratically with sequence length due to full self-attention.

Leveraging Pre-trained Language Model Checkpoints for Encoder-Decoder Models

Breaking the Likelihood Trap: Variance-Calibrated Modulation for Large Language Model Decoding

The paper introduces Variance‑Calibrated Modulation (VCM), a training‑free pre‑decoding technique that reshapes language model probability distributions before truncation. VCM uses two dynamic mechanisms: a Contextual Searchlight via PMI to suppress stopwords and highlight context‑relevant tokens, and an Adaptive Self‑Debiasing that applies scale‑invariant penalization based on real‑time logit standard deviation. Experiments on open‑ended generation, factual QA, and mathematical reasoning show that VCM consistently reduces the likelihood trap, improving diversity, coherence, and reasoning accuracy with minimal computational cost.

Dynamic-dLLM: Dynamic Cache-Budget and Adaptive Parallel Decoding for Training-Free Acceleration of Diffusion LLM

arXiv:2606. 26120v1 Announce Type: cross Abstract: Diffusion Large Language Models (dLLMs) offer a promising alternative to autoregressive models, excelling in text generation tasks due to their bidirectional attention mechanisms.

Adaptive Multi-Step Lookahead Decoding for Diffusion Language Models

arXiv:2607. 15655v1 Announce Type: cross Abstract: Masked diffusion language models (DLMs) enable parallel text generation by iteratively refining masked tokens, offering a promising alternative to autoregressive decoding.

Orchestrating Dual-Boundaries: An Arithmetic Intensity Inspired Acceleration Framework for Diffusion Language Models

arXiv:2511. 21759v2 Announce Type: replace-cross Abstract: Diffusion-based large language models (dLLMs) have recently gained significant attention for their exceptional performance and inherent potential for parallel decoding.

Controllably Efficient Language Models

arXiv:2511. 05313v2 Announce Type: replace Abstract: The substantial inference costs of attention in transformers motivated the development of efficient sequence mixers: namely sparse and sliding window attention, convolutions and linear attention.