ResearchResearch paperSpeech & Audio · Large Language Models · Training & Scaling1 source · Oct 7, 2026

A Strength-Monotonic Law for Domain Alignment in Frozen-Embedding Bioacoustic Classification

When does distribution alignment help a frozen foundation-model embedding generalize across acoustic domains?

For cross-domain mosquito-species classification we report a strength-monotonic law: the stronger an encoder is on the target task, the more its unseen-domain generalization relies on a distribution-alignment (MMD) term, and the more it is harmed by domain-rebalanced sampling.
Across four encoder families and a within-encoder HuBERT layer sweep (n=8), the rebalancing leg orders exactly with encoder strength (Spearman -1.000), while the MMD-benefit leg is monotonic within each stream and -0.857 pooled; fixing architecture and varying only representation strength flips the rebalancing effect from benefit to collapse.
The law is actionable: a single MMD term is the sole lever on a strong encoder, so we reduce the field's default recipe to a frozen Perch 2.0 embedding, a lightweight probe, cross-entropy, one MMD, and input augmentation.
As boundary conditions of the same law, three community defaults (backbone fine-tuning, multi-modal fusion, and domain rebalancing) each hurt unseen-domain accuracy under a leave-domain protocol, shown with single-variable, multi-seed evidence.

[1]A Strength-Monotonic Law for Domain Alignment in Frozen-Embedding Bioacoustic Classification
arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 09:31 AM
When does distribution alignment help a frozen foundation-model embedding generalize across acoustic domains?
For cross-domain mosquito-species classification we report a strength-monotonic law: the stronger an encoder is on the target task, the more its unseen-domain generalization relies on a distribution-alignment (MMD) term, and the more it is harmed by domain-rebalanced sampling.

Extractive summary: sentences quoted from the sources.

Before this