SelfJudge: Faster Speculative Decoding via Self-Supervised Judge Verification

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Yoon, Kanghoon, Kim, Minsub, Lee, Sungjae, Lee, Joonhyung, Woo, Sunghyeon, In, Yeonjun, Kwon, Se Jung, Park, Chanyoung, Lee, Dongsoo
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917539319119872
author Yoon, Kanghoon
Kim, Minsub
Lee, Sungjae
Lee, Joonhyung
Woo, Sunghyeon
In, Yeonjun
Kwon, Se Jung
Park, Chanyoung
Lee, Dongsoo
author_facet Yoon, Kanghoon
Kim, Minsub
Lee, Sungjae
Lee, Joonhyung
Woo, Sunghyeon
In, Yeonjun
Kwon, Se Jung
Park, Chanyoung
Lee, Dongsoo
contents Speculative decoding accelerates LLM inference by verifying candidate tokens from a draft model against a larger target model. Recent judge decoding boosts this process by relaxing verification criteria by accepting draft tokens that may exhibit minor discrepancies from target model output, but existing methods are restricted by their reliance on human annotations or tasks with verifiable ground truths, limiting generalizability across diverse NLP tasks. We propose SelfJudge, which trains judge verifiers via self-supervision of the target model. Our method measures semantic preservation by assessing whether token-substituted responses preserve the meaning of original responses, enabling automatic verifier training across diverse NLP tasks. Our experiments show SelfJudge achieves superior inference-accuracy trade-offs than judge decoding baselines, offering a broadly applicable solution for faster LLM inference.
format Preprint
id arxiv_https___arxiv_org_abs_2510_02329
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SelfJudge: Faster Speculative Decoding via Self-Supervised Judge Verification
Yoon, Kanghoon
Kim, Minsub
Lee, Sungjae
Lee, Joonhyung
Woo, Sunghyeon
In, Yeonjun
Kwon, Se Jung
Park, Chanyoung
Lee, Dongsoo
Computation and Language
Artificial Intelligence
Speculative decoding accelerates LLM inference by verifying candidate tokens from a draft model against a larger target model. Recent judge decoding boosts this process by relaxing verification criteria by accepting draft tokens that may exhibit minor discrepancies from target model output, but existing methods are restricted by their reliance on human annotations or tasks with verifiable ground truths, limiting generalizability across diverse NLP tasks. We propose SelfJudge, which trains judge verifiers via self-supervision of the target model. Our method measures semantic preservation by assessing whether token-substituted responses preserve the meaning of original responses, enabling automatic verifier training across diverse NLP tasks. Our experiments show SelfJudge achieves superior inference-accuracy trade-offs than judge decoding baselines, offering a broadly applicable solution for faster LLM inference.
title SelfJudge: Faster Speculative Decoding via Self-Supervised Judge Verification
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2510.02329