RetroThinker: Enabling Retrospective Thinking in Speech LLMs

2026-09-10

AI Summary

本文通过引入RetroThinker框架，采用监督微调和直接偏好优化方法，解决了语音大模型在复杂推理任务中准确性和延迟之间的权衡问题。

Abstract

Speech large language models (SpeechLLMs) offer reduced latency and retain paralinguistic nuances that are typically lost in cascaded automatic speech recognition (ASR) and text-based LM architectures. However, they continue to lag behind text-only LLMs on complex reasoning tasks, while real-time spoken interaction imposes strict latency constraints. Although prior works employ Chain-of-Thought (CoT) and concurrent reasoning to enhance reasoning capabilities without inducing prohibitive delays, an inherent accuracy-latency trade-off persists. In this paper, we investigate whether a streaming SpeechLLM can dynamically revise its reasoning traces on the fly. We introduce RetroThinker, a multi-stage post-training framework that equips the Moshi model to self-verify and forward-correct CoT steps during inference. RetroThinker combines supervised fine-tuning (SFT) on curated retrospective thinking data with length-based direct preference optimization (DPO) to optimize retrospective during early reasoning (i.e., reasoning concurrently while the user speaks). Evaluated on the GSM8K benchmark, RetroThinker significantly improves the accuracy-latency trade-off over non-retrospective baselines, achieving an 11% absolute accuracy gain at a comparable latency.

Problem

• Speech Large Language Models
• Complex Reasoning Tasks
• Latency Constraints
• Accuracy-Latency Trade-off
• Chain-of-Thought

Innovation

• RetroThinker
• supervised fine-tuning (SFT)
• direct preference optimization (DPO)
• Chain-of-Thought (CoT)
• real-time spoken interaction