Beyond the Target: From Imitation to Collaboration in Speculative Decoding 事件

Beyond the Target: From Imitation to Collaboration in Speculative Decoding arXiv:2605.24793v1 Announce Type: new Abstract: Speculative decoding (SPD) accelerates large language model (LLM) inference by letting a smaller draft model propose multiple future tokens that are verified in parallel by a larger target model. The dominant SPD paradigm treats the target model as the sole reliable teacher, accepting a draft token only when it exactly matches the target prediction. This design implicitly a