# CoReLoop: Parameter-Efficient Controlled Recurrent Refinement for Audio Deepfake Detection

arXiv: https://arxiv.org/abs/2609.19818 · HTML: https://mlsanity.com/paper/2609.19818/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 24 references that are in the map.
- First posted on arXiv 17 Sep 2026.

## Builds on (most cited first, 21 of 21)

- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach](https://mlsanity.com/paper/2502.05171/) (2025)
- [GigaSpeech: An Evolving, Multi-domain ASR Corpus with 10,000 Hours of Transcribed Audio](https://mlsanity.com/paper/2106.06909/) (2021)
- [Automatic speaker verification spoofing and deepfake detection using wav2vec 2.0 and data augmentation](https://mlsanity.com/paper/2202.12233/) (2022)
- [AISHELL-3: A Multi-speaker Mandarin TTS Corpus and the Baselines](https://mlsanity.com/paper/2010.11567/) (2020)
- [ADD 2022: the First Audio Deep Synthesis Detection Challenge](https://mlsanity.com/paper/2202.08433/) (2022)
- [ASVspoof 5: Crowdsourced Speech Data, Deepfakes, and Adversarial Attacks at Scale](https://mlsanity.com/paper/2408.08739/) (2024)
- [ADD 2023: the Second Audio Deepfake Detection Challenge](https://mlsanity.com/paper/2305.13774/) (2023)
- [One-class Learning Towards Synthetic Voice Spoofing Detection](https://mlsanity.com/paper/2010.13995/) (2020)
- [MLAAD: The Multi-Language Audio Anti-Spoofing Dataset](https://mlsanity.com/paper/2401.09512/) (2024)
- [The Codecfake Dataset and Countermeasures for the Universally Detection of Deepfake Audio](https://mlsanity.com/paper/2405.04880/) (2024)
- [CN-Celeb: multi-genre speaker recognition](https://mlsanity.com/paper/2012.12468/) (2020)
- [SpoofCeleb: Speech Deepfake Detection and SASV In The Wild](https://mlsanity.com/paper/2409.17285/) (2024)
- [Speech DF Arena: A Leaderboard for Speech DeepFake Detection Models](https://mlsanity.com/paper/2509.02859/) (2025)
- [DFADD: The Diffusion and Flow-Matching Based Audio Deepfake Dataset](https://mlsanity.com/paper/2409.08731/) (2024)
- [Harder or Different? Understanding Generalization of Audio Deepfake Detection](https://mlsanity.com/paper/2406.03512/) (2024)
- [SpeechFake: A Large-Scale Multilingual Speech Deepfake Dataset Incorporating Cutting-Edge Generation Methods](https://mlsanity.com/paper/2507.21463/) (2025)
- [RecurTrace: Adaptive Latent Reasoning with Loop-Time Memory](https://mlsanity.com/paper/2609.03379/) (2026)
- [Generalizable speech deepfake detection via meta-learned LoRA](https://mlsanity.com/paper/2502.10838/) (2025)
- [Teffic-Audio: Tell Fact from Fiction](https://mlsanity.com/paper/2607.28351/) (2026)
- [REIMU: Efficient Heterogeneous Hierarchical Reasoning for SSL-Based Speech Deepfake Detection](https://mlsanity.com/paper/2608.00857/) (2026)

## Built on by (most cited first, 2 of 2)

- [AdaLoop: Adaptive-Depth Latent Reasoning for Audio Language Models](https://mlsanity.com/paper/2610.06949/) (2026)
- [SkillFormer: Skill-Decomposed Adaptation for Audio Language Models](https://mlsanity.com/paper/2610.07533/) (2026)

## Abstract (arXiv)

Audio deepfake detectors are expected to remain robust to diverse attacks, datasets, and recording conditions, yet collecting training data that cover every future condition is impractical. We investigate whether an already-trained SSL-based detector can improve its generalization through additional recurrent computation, without introducing new training data or updating its original parameters. Directly feeding the encoder output back as its next input, however, severely degrades detection, revealing a mismatch between the representations produced and consumed by a model trained for single-pass inference. We therefore propose CoReLoop, which enables controlled recurrent refinement through loop-specific LoRA and lightweight modules that adapt recurrent inputs, regulate state updates, and align refined representations with the frozen classifier. Across 14 diverse test sets, CoReLoop improves pooled EER across 3-, 6-, 12-, and 24-layer encoders; on the 24-layer model, two-pass refinement reduces EER from 4.85% to 3.74% while training only approximately 11M of 598M parameters. An optional halting head further selects refinement depth per utterance, achieving 3.73% EER with an average of 1.18 passes. Analysis shows that later-pass saturation coincides with smaller applied updates and nearly balanced corrections and regressions, while the recurrent state continues to change.