# Reject Only Critical Tokens: Pivot-Aware Speculative Decoding

arXiv: https://arxiv.org/abs/2511.00351 · HTML: https://mlsanity.com/paper/2511.00351/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Speculative decoding, one of 12 directions in Efficient LLM inference and serving.
- Builds on 17 of its 22 references that are in the map.
- First posted on arXiv 1 Nov 2025.

## Placed in

- Efficient LLM inference and serving / [Speculative decoding](https://mlsanity.com/inference/speculative-decoding/)

## Builds on (most cited first, 17 of 17)

- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Fast Inference from Transformers via Speculative Decoding](https://mlsanity.com/paper/2211.17192/) (2022)
- [Accelerating Large Language Model Decoding with Speculative Sampling](https://mlsanity.com/paper/2302.01318/) (2023)
- [Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads](https://mlsanity.com/paper/2401.10774/) (2024)
- [EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty](https://mlsanity.com/paper/2401.15077/) (2024)
- [Speculative Decoding with Big Little Decoder](https://mlsanity.com/paper/2302.07863/) (2023)
- [REST: Retrieval-Based Speculative Decoding](https://mlsanity.com/paper/2311.08252/) (2023)
- [Reward-Guided Speculative Decoding for Efficient LLM Reasoning](https://mlsanity.com/paper/2501.19324/) (2025)
- [PEARL: Parallel Speculative Decoding with Adaptive Draft Length](https://mlsanity.com/paper/2408.11850/) (2024)
- [Cascade Speculative Drafting for Even Faster LLM Inference](https://mlsanity.com/paper/2312.11462/) (2023)
- [GliDe with a CaPE: A Low-Hassle Method to Accelerate Speculative Decoding](https://mlsanity.com/paper/2402.02082/) (2024)
- [Judge Decoding: Faster Speculative Sampling Requires Going Beyond Model Alignment](https://mlsanity.com/paper/2501.19309/) (2025)
- [SpecReason: Fast and Accurate Inference-Time Compute via Speculative Reasoning](https://mlsanity.com/paper/2504.07891/) (2025)
- [Ouroboros: Generating Longer Drafts Phrase by Phrase for Faster Speculative Decoding](https://mlsanity.com/paper/2402.13720/) (2024)
- [Scaling Speculative Decoding with Lookahead Reasoning](https://mlsanity.com/paper/2506.19830/) (2025)

## Built on by (most cited first, 4 of 4)

- [JuDi: Revisiting Judge Decoding from First Principles via Training-Free Distributional Divergence](https://mlsanity.com/paper/2601.04766/) (2026)
- [Beyond the Target: From Imitation to Collaboration in Speculative Decoding](https://mlsanity.com/paper/2605.24793/) (2026)
- [ASPIRE: Asynchronous Batched Self-Speculative Decoding for Long-Context LLM Inference](https://mlsanity.com/paper/2609.17943/) (2026)
- [A Calibrated Instrument for Measuring How Inference Optimizations Affect Output Quality](https://mlsanity.com/paper/2609.18005/) (2026)

## Abstract (arXiv)

Speculative Decoding (SD) ensures that the output matches the target model's distribution exactly. However, we argue that this distribution matching requirement is too stringent and results in unnecessarily low acceptance rates, limiting potential speedups. Instead, we advocate a reformulation of the decoding objective: the proposed decoding strategy should match the expected utility, i.e., the task-specific performance, of the target model. This perspective also aligns better with real-world use cases of LLMs, where utility (e.g., code correctness, factual accuracy) is often more important than sampling distribution. Based on this reformulation, we propose a novel decoding strategy: Pivot-Aware Speculative Decoding, which rejects only those tokens that would lead to a utility drop in the final output. We refer to these critical tokens as pivot tokens. We propose a method for labeling tokens as pivotal or non-pivotal and train a lightweight classifier to detect them. This method can be viewed as a relaxed version of standard SD, which offers much higher acceptance while preserving utility. We evaluate our method across various datasets, demonstrating that we can achieve up to 2.5× speedup with comparable utility. Source code is available at https://github.com/amir-zsh/PAD.