Efficient Training-Free Multi-Token Prediction via Embedding-Space Probing 文章

ArXiv CS.CL2026-05-29NEWSen作者: Raghavv Goel, Mukul Gagrani, Mingu Lee, Chris Lott

摘要

arXiv:2603.17942v2 Announce Type: replace Abstract: Large Language Models (LLMs) possess latent multi-token prediction (MTP) abilities despite being trained only for next-token generation. We introduce ESP (Embedding-Space Probing), a simple and training-free MTP method that probes an LLM using on-the-fly mask tokens drawn from its embedding space, enabling parallel future-token prediction without modifying weights or relying on draft models. ESP constructs a speculative token tree by sampling Top-K candidates from mask-token logits and applies a lightweight pruning rule to retain high-probability continuations. During generation, predictions are verified in parallel, yielding lossless decoding while significantly reducing model calls and increasing token throughput. ESP consistently outperforms existing training-free baselines, improving acceptance length by 7-11% over LADE on LLaMA3 and 7-8% on Qwen3, and increasing throughput by up to 15-19% over the strongest baseline.

相关公司

暂无数据

相关人物

暂无数据