SpecSteer: Synergizing Local Context and Global Reasoning for Efficient Personalized Generation

TL;DR

SpecSteer is a collaborative inference framework that combines on-device personalization with cloud reasoning, enhancing personalized generation while preserving privacy and achieving significant speed improvements.

Contribution

It introduces a novel asymmetric collaboration method using Bayesian knowledge fusion and speculative decoding for privacy-preserving personalized generation.

Findings

Achieves 2.36x speedup over standard baselines.

Successfully bridges the reasoning gap in personalized generation.

Demonstrates superior performance in experiments.

Abstract

Realizing personalized intelligence faces a core dilemma: sending user history to centralized large language models raises privacy concerns, while on-device small language models lack the reasoning capacity required for high-quality generation. Our pilot study shows that purely local enhancements remain insufficient to reliably bridge this gap. We therefore propose SpecSteer, an asymmetric collaborative inference framework that synergizes private on-device context with cloud-scale reasoning. SpecSteer casts collaboration as Bayesian knowledge fusion and repurposes speculative decoding as a distributed alignment protocol, yielding a Draft--Verify--Recover pipeline: the on-device model drafts personalized sequences; the cloud validates via a ratio-based mechanism that decouples reasoning verification from private context, filtering logical flaws without accessing raw user context; upon rejection, a steering recovery injects local intent during correction. Experiments demonstrate that SpecSteer successfully closes the reasoning gap and achieves superior personalized generation performance, while delivering a 2.36x speedup over standard baselines.