פיענוח מונחה: אינפרנס מהיר פוגש תגבור

תקציר מקורי באנגליתarXiv:2609.30474v1 Announce Type: new Abstract: Speculative decoding is a successful technique speeding up inference of a target autoregressive language model via a fast drafter model. Lossy speculative decoding allows a drift with respect to the target to further improve speed. Interestingly, it has been observed experimentally that the resulting model can $\textit{also}$ beat the target $\textit{quality-wise}$. Our paper formally proves how such a feat is possible with a formal approach to lossy speculative decoding called $\textit{mentored decoding}$. To get there, we connect inference to a celebrated ML training theory, $\textit{boosting}$, and proceed via the generalization of mentored decoding to the whole set of $f$-divergences. We uncover key properties of mentored decoding, among