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

arXiv: https://arxiv.org/abs/2610.09737 · HTML: https://mlsanity.com/paper/2610.09737/

## Key facts

- No citing papers in the map yet (data through 7 Oct 2026).
- Builds on 27 of its 30 references that are in the map.
- First posted on arXiv 7 Oct 2026.

## Builds on (most cited first, 27 of 27)

- [Supervised Contrastive Learning](https://mlsanity.com/paper/2004.11362/) (2020)
- [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://mlsanity.com/paper/2106.07447/) (2021)
- [Learning Transferable Features with Deep Adaptation Networks](https://mlsanity.com/paper/1502.02791/) (2015)
- [Deep CORAL: Correlation Alignment for Deep Domain Adaptation](https://mlsanity.com/paper/1607.01719/) (2016)
- [Robust fine-tuning of zero-shot models](https://mlsanity.com/paper/2109.01903/) (2021)
- [Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution](https://mlsanity.com/paper/2202.10054/) (2022)
- [Deep Domain Confusion: Maximizing for Domain Invariance](https://mlsanity.com/paper/1412.3474/) (2014)
- [BEATs: Audio Pre-Training with Acoustic Tokenizers](https://mlsanity.com/paper/2212.09058/) (2022)
- [Surgical Fine-Tuning Improves Adaptation to Distribution Shifts](https://mlsanity.com/paper/2210.11466/) (2022)
- [AVES: Animal Vocalization Encoder based on Self-Supervision](https://mlsanity.com/paper/2210.14493/) (2022)
- [BirdSet: A Large-Scale Dataset for Audio Classification in Avian Bioacoustics](https://mlsanity.com/paper/2403.10380/) (2024)
- [NatureLM-audio: an Audio-Language Foundation Model for Bioacoustics](https://mlsanity.com/paper/2411.07186/) (2024)
- [Can Masked Autoencoders Also Listen to Birds?](https://mlsanity.com/paper/2504.12880/) (2025)
- [Perch 2.0: The Bittern Lesson for Bioacoustics](https://mlsanity.com/paper/2508.04665/) (2025)
- [Transferable Models for Bioacoustics with Human Language Supervision](https://mlsanity.com/paper/2308.04978/) (2023)
- [BIRB: A Generalization Benchmark for Information Retrieval in Bioacoustics](https://mlsanity.com/paper/2312.07439/) (2023)
- [Foundation Models for Bioacoustics -- a Comparative Review](https://mlsanity.com/paper/2508.01277/) (2025)
- [Device-Robust Acoustic Scene Classification via Impulse Response Augmentation](https://mlsanity.com/paper/2305.07499/) (2023)
- [Mind the Domain Gap: a Systematic Analysis on Bioacoustic Sound Event Detection](https://mlsanity.com/paper/2403.18638/) (2024)
- [Leveraging tropical reef, bird and unrelated sounds for superior transfer learning in marine bioacoustics](https://mlsanity.com/paper/2404.16436/) (2024)
- [Data-Efficient Low-Complexity Acoustic Scene Classification in the DCASE 2024 Challenge](https://mlsanity.com/paper/2405.10018/) (2024)
- [AVEX: What Matters for Animal Vocalization Encoding](https://mlsanity.com/paper/2508.11845/) (2025)
- [Linear probing enables Ship-Radiated Noise recognition with pretrained audio embeddings](https://mlsanity.com/paper/2601.08358/) (2026)
- [Feature-Weighted MMD-CORAL for Domain Adaptation in Power Transformer Fault Diagnosis](https://mlsanity.com/paper/2505.14896/) (2025)
- [Learning Domain-Robust Bioacoustic Representations for Mosquito Species Classification with Contrastive Learning and Distribution Alignment](https://mlsanity.com/paper/2510.00346/) (2025)
- [BioDCASE 2026 Challenge Baseline for Cross-Domain Mosquito Species Classification](https://mlsanity.com/paper/2603.20118/) (2026)
- [Is Domain Adaptation Always Helpful? A Frozen-Backbone Study of Cross-Domain Sentiment Transfer](https://mlsanity.com/paper/2607.05937/) (2026)

## Abstract (arXiv)

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. The reduced recipe stays within seed noise of the full composite (BA_unseen 0.299+/-0.006 vs. 0.307+/-0.014). 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. We present a mechanism and the recipe it explains, not a leaderboard entry.