Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Enhancing Speech Large Language Models through Reinforced Behavior Alignment

About

The recent advancements of Large Language Models (LLMs) have spurred considerable research interest in extending their linguistic capabilities beyond text to other modalities, which leads to emergence of speech-based LLMs (SpeechLMs) with capability of processing user request in either speech or textual formats. However, owing to inter-modal discrepancies, these SpeechLMs still exhibit a significant performance gap compared to their text-based LLM counterparts in instruction-following, particularly when confronted with the dynamic and variable nature of user speech. To address this challenge, this paper introduces a framework termed Reinforced Behavior Alignment (RBA), designed to bolster the language generation proficiency of SpeechLMs. Instead of relying on supervised fine-tuning from human annotations, RBA employs a self-synthesis methodology to generate extensive, high-fidelity alignment data by a powerful teacher LLM. Then SpeechLMs is aligned its behavior with that of a teacher using a reinforcement learning-based approach. Experimental results demonstrate that this method effectively enhances the instruction-following capabilities of SpeechLMs that outperform conventional distillation baselines. Crucially, we demonstrate that RBA can be seamlessly extended to tasks such including spoken question answering and speech-to-text translation, attaining state-of-the-art performance on open benchmarks with only self-generated data.

Yansong Liu, Jiateng Li, Yuan Liu• 2025

Related benchmarks

TaskDatasetResultRank
Speech-to-Speech Question-AnsweringWebQ
Accuracy51.3
25
Instruction FollowingSpeech-IFEval
IF Rate68.2
18
Spoken Question AnsweringTriviaQA
Accuracy65.2
15
Speech-to-text TranslationFLEURS (test)
Avg. Score36.8
15
Instruction FollowingSpoken-Alpaca
LC Score79.4
11
Speech robustnessDOWIS
LC70.1
11
Speech robustnessVocalBench-DF
LC Score65
11
Speech robustnessSpeech-view Consistency
Consistency96.9
11
Speech-to-text TranslationEnglish-to-German en-de
BLEU Score34
10
Speech-to-text TranslationEnglish-to-Chinese (En->Zh)
BLEU Score49.1
10
Showing 10 of 22 rows

Other info

Follow for update