Think Before You Transcribe: Context-Steered Reasoning for Hotword-Aware Speech Recognition

Abstract

Contextual automatic speech recognition (ASR) improves rare-entity recognition by supplying candidate hotwords, but retrieved candidates may be incomplete or incorrect. A reliable system must therefore exploit a relevant candidate without allowing unsupported alternatives to override the acoustic evidence. We propose a context-steered reasoning framework that generates a short, audio-grounded caption before transcription. This task-specific representation lets the model assess candidate relevance before making a lexical decision. Supervised fine-tuning establishes the structured caption-to-transcription behavior, while targeted trajectory injection strengthens group relative policy optimization when difficult utterances yield only unsuccessful rollouts. Experiments across diverse context conditions show that the proposed method consistently improves both transcription quality and hotword recognition while remaining robust to irrelevant candidates.

The authors’ abstract, as arXiv lists it. arXiv’s listings, abstracts included, are free for anyone to reuse.

More in Computation & Language / NLP

• arXiv preprint · 19 October 2026

  Cognitive Maps in Large Language Models

  Albert Qin, Jonathan Cohen and Gautam Reddy
• arXiv preprint · 20 October 2026

  The Router Already Knows

  Andrew Bistras and Marty Johnson
• arXiv preprint · 19 October 2026

  Text Dominance in Paralinguistic Conflict: Tracing Suppression Through Audio LLM Circuits

  Ankur Banerjee, Vishnu Guntakala, Danity Pike and 3 others
• arXiv preprint · 20 October 2026

  Harness Self-Evolution for Long-Horizon Terminal Agents

  Ruhan Wang, Yucheng Shi, Zongxia Li and 6 others
• arXiv preprint · 19 October 2026

  Curriculum Learning Teaches LLMs Long-Horizon Implicit Reasoning, Up To A Limit

  Jimson Huang, Maxwell Nickell, Ruowang Zhang and 1 other

All new papers in Computation & Language / NLP