Hugging Face Trending Papers

Windowed-MTP: Removing the Full-Context Draft-KV Tax at Million-Token Context

Read the original on Hugging Face Trending Papers →

Speculative decoding accelerates autoregressive generation by having a cheap draft propose tokens that a target verifies in parallel. Frontier models increasingly ship a built-in Multi-Token-Prediction (MTP/NEXTN) draft head under the assumption that the draft is negligibly cheap.

Summary generated by The Flow from the publisher's feed. The full article lives at Hugging Face Trending Papers.

Hugging Face Trending Papers
Jul 27

DraftExpert: Expansion-Aware Self-Speculative Decoding for End-Device MoE Inference

Large Mixture-of-Experts (MoE) language models are attractive for end-device deployment because only a small subset of experts is active per token, but their routed expert weights often exceed accelerator memory. We target latency-critical single-user settings where routed experts are staged on demand from CPU memory to a GPU or from Flash to a mobile NPU.

arXiv Machine Learning
Jul 20

DualKV: Shared-Prompt Flash Attention for Efficient RL Training with Large Rollouts and Long Contexts

arXiv:2605. 15422v3 Announce Type: replace Abstract: Modern RL post-training methods such as GRPO and DAPO train on N response sequences of R tokens sampled from a shared prompt of P tokens, but standard FlashAttention replicates all P prompt tokens N times across both forward and backward passes -- duplicating compute and memory on identical hidden states.

By Jiading Gai, Shuai Zhang, Xiang Song, Bernie Wang, George Karypis
arXiv AI
Jun 29

Bifocal Diffusion Language Models: Asymmetric Bidirectional Context for Parallel Generation

arXiv:2606. 27732v1 Announce Type: cross Abstract: Discrete diffusion language models (dLLMs) recover masked tokens in parallel, offering significant speedups over autoregressive (AR) generation.

By Yuhang Chen, Xianfeng Wu, Jinhao Duan, Mingfu Liang, Xiaohan Wei, Yunchen Pu, Fei Tian, Chonglin Sun, Parish Aggarwal, Frank Shyu, Luke Simon, Sandeep Pandey, Xi Liu, Tianlong Chen