Match the Distribution, Not the Compute: Post-Training Multi-Token Prediction Heads

arXiv:2610.00888v1 Announce Type: cross Abstract: Multi-token prediction (MTP) improves the throughput of autoregressive generation by enabling the language model to draft multiple next tokens per forward pass, while a verification step over draft tokens ensures that token distribution of the backbone is preserved. Every open MTP-family release (MiMo-7B, DeepSeek-V3, Qwen3) trains its heads jointly with the backbone over the full pretraining run of tens of trillions of tokens, thus setting the drafter quality at pretraining time. We ask whether a lightweight post-training pass on target-generated chain-of-thought is enough to reach the same expected throughput speedup on a frozen reasoning model, and study how a serving-time system built on such a checkpoint can be optimized. We present three findings. 1) On a frozen Qwen3-8B with K{=}3 chained MTP heads, we show that a post-training recipe with plain cross-entropy on approx!2.5B tokens reaches or exceeds the expected speedup of jointly trained MiMo-7B on math, coding and knowledge benchmarks. Our post-training recipe utilizes 10^3-10^4imes less MTP-training tokens as compared with joint pre-training of MiMO-7B MTP baseline. 2) We propose a chain-aware relaxation of draft token verification rule that allows a bounded drift from backbone language model token distribution. We show that this relaxation lifts expected speedups by +12 to +16% per benchmark while preserving task accuracy. 3) We propose an adaptive controller that dynamically chooses the number of MTP heads to be engaged at inference time and demonstrate recovery of upto 11--14% loss in speedup using fixed maximum MTP draft length.

Related

• Free Pause Tokens
• CLP: Collocation-Length Prediction for Zero-Loss Adaptive Multi-Token Inference
• SMART: When is it Actually Worth Expanding a Speculative Tree?
• A Practical Investigation of Training-free Relaxed Speculative Decoding

Source: arXiv cs.AI | 2026-10-02