General-purpose audio representations aim to map acoustically variable instances of the same event to nearby points, resolving content identity in a zero-shot setting. Unlike supervised classification benchmarks that measure adaptability via parameter updates, we introduce VocSim, a training-free benchmark probing the intrinsic geometric alignment of frozen embeddings, with no parameters updated and no labels used (a label-free PCA whitening is fit per subset to correct anisotropy). VocSim aggregates 125k single-source clips from 19 corpora spanning human speech, animal vocalizations, and environmental sounds, isolating content representation from source separation (polyphonic mixtures are out of scope). We evaluate embeddings with Precision@k for local purity and the Global Separation Rate (GSR) for point-wise class separation, calibrated by lift over an empirical permutation baseline. A simple pipeline of frozen Whisper features, time-frequency pooling, and label-free PCA yields strong zero-shot performance with stable GSR rankings across domains (Kendall's tau = 0.60). However, on blind low-resource speech (Shipibo-Conibo, Chintang), local retrieval collapses while remaining above chance, exposing a cross-lingual speech generalization gap. As external validation, our top embeddings predict avian perceptual similarity, improve bioacoustic classification, and achieve state-of-the-art on the HEAR benchmark. We release data, code, and a public leaderboard.

翻译：通用音频表示旨在将同一事件的不同声学变体实例映射到邻近点，从而在零样本设置下解决内容身份识别问题。与通过参数更新衡量适应性的有监督分类基准不同，我们提出VocSim——一个无需训练即可探查冻结嵌入内在几何对齐的基准（不进行参数更新，不使用标签，仅对每个子集拟合无标签PCA白化以校正各向异性）。VocSim整合了来自19个语料库的12.5万个单源音频片段，涵盖人类语音、动物发声和环境声音，并将内容表征与源分离问题隔离（多声部混合音频不在研究范围内）。我们使用局部纯净度的Precision@k和基于经验排列基线提升校准的点态类别分离全局分离率（GSR）来评估嵌入。一个由冻结Whisper特征、时频池化和无标签PCA组成的简单流水线在零样本任务中表现强劲，且GSR排名在跨域场景下保持稳定（肯德尔τ系数=0.60）。然而，在低资源盲语音场景（Shipibo-Conibo语、Chintang语）中，局部检索性能虽高于随机水平但显著下降，暴露了跨语言语音泛化差距。作为外部验证，我们的顶级嵌入能够预测鸟类感知相似度、提升生物声学分类效果，并在HEAR基准上达到当前最优水平。我们已公开数据集、代码和公共排行榜。