Wenxiang Guo

ProactiveAudioBench: Evaluating Condition-Driven Responses in Audio Streams

Audio assistants are moving from turn-based question answering toward continuous listening, where they must interpret incoming audio while deciding whether to respond or remain silent. Existing proactive benchmarks emphasize video and audiovisual interaction, while audio-centric...

LAMAR-Bench: Evaluating Long-Form and Multi-Audio Understanding and Reasoning

Audio-language models increasingly process long recordings and multiple audio inputs, requiring fine-grained acoustic understanding, cross-audio reasoning, and decisions under goals and constraints. Existing benchmarks cover these requirements unevenly, while answer accuracy can...