arXiv：cs.AI· Hongyu Jin, Siyi Wang, Yang Xiao, Jiaheng Dong, Shihong Tan, Kaiyuan Peng, Georgiana Juravle, Shanquan Chen, Gongping Huang, Hong Jia, Eun-Jung Holden, James Bailey, Ting Dang·· 5 小时前
RAIL：基于 CHC 框架重新审视大型音频语言模型的听觉智能评测
RAIL: Rethinking Auditory Intelligence in Large Audio-Language Models with a CHC-Grounded Benchmark
AI 导读
研究者提出 RAIL，一个以 Cattell-Horn-Carroll（CHC）框架为基础、面向人类的听觉认知基准，将听觉认知形式化为五项核心能力并构建结构化评测任务。
正文
Authors:Hongyu Jin, Siyi Wang, Yang Xiao, Jiaheng Dong, Shihong Tan, Kaiyuan Peng, Georgiana Juravle, Shanquan Chen, Gongping Huang, Hong Jia, Eun-Jung Holden, James Bailey, Ting Dang
Abstract:Humans process rich auditory environments through various cognitive capabilities such as audio perception, audio reasoning and memory. Despite recent progress in large audio-language models (LALMs), current evaluation remains largely task- or domain-centric, focusing on end performance while ignoring underlying auditory cognitive behaviors. This gap between how auditory cognition is understood in humans and how it is evaluated in LALMs hinders the interpretation of model behaviour in underlying capabilities and limits alignment with human auditory objectives. We introduce RAIL, a human-centric benchmark grounded in the Cattell--Horn--Carroll (CHC) framework, which formalises auditory cognition into five core capabilities, and develops them into structured evaluation tasks with principled data curation and human-aligned evaluation protocols. Evaluating 26 LALMs reveals strong performance in knowledge-related tasks but weak auditory perception and memory, reflecting reliance on language-based pretraining. Limited reasoning under auditory settings further suggests a mismatch between existing training paradigms and human-like auditory reasoning, as well as over-reasoning at the cost of efficiency. Six LALMs exceed human performance in general, while auditory processing all lags behind human. Overall, RAIL establishes a new framework for human-aligned evaluation of auditory intelligence.
| Comments: | Accepted as a poster at NeurIPS 2026. Code: this https URL ; Dataset: this https URL ; Project page: this https URL |
| Subjects: | Sound (cs.SD); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2606.11260 [cs.SD] |
| (or arXiv:2606.11260v2 [cs.SD] for this version) | |
| https://doi.org/10.48550/arXiv.2606.11260 arXiv-issued DOI via DataCite |
Submission history
From: Hongyu Jin [view email]
[v1]
Tue, 9 Jun 2026 02:38:17 UTC (9,608 KB)
[v2]
Sun, 11 Oct 2026 00:42:42 UTC (9,608 KB)
来源：arXiv：cs.AI · arxiv.org