# Subliminal Corruption: Mechanisms, Thresholds, and Interpretability

arXiv: https://arxiv.org/abs/2510.19152 · HTML: https://mlsanity.com/paper/2510.19152/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 17 references that are in the map.
- First posted on arXiv 22 Oct 2025.

## Builds on (most cited first, 8 of 8)

- [Concrete Problems in AI Safety](https://mlsanity.com/paper/1606.06565/) (2016)
- [Discovering Latent Knowledge in Language Models Without Supervision](https://mlsanity.com/paper/2212.03827/) (2022)
- [RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback](https://mlsanity.com/paper/2309.00267/) (2023)
- [Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training](https://mlsanity.com/paper/2401.05566/) (2024)
- [Alignment faking in large language models](https://mlsanity.com/paper/2412.14093/) (2024)
- [Simple synthetic data reduces sycophancy in large language models](https://mlsanity.com/paper/2308.03958/) (2023)
- [What is in Your Safe Data? Identifying Benign Data that Breaks Safety](https://mlsanity.com/paper/2404.01099/) (2024)
- [Subliminal Learning: Language models transmit behavioral traits via hidden signals in data](https://mlsanity.com/paper/2507.14805/) (2025)

## Built on by (most cited first, 4 of 4)

- [From Data to Behavior: Predicting Unintended Model Behaviors Before Training](https://mlsanity.com/paper/2602.04735/) (2026)
- [Emergent and Subliminal Misalignment Through the Lens of Data-Mediated Transfer](https://mlsanity.com/paper/2605.12798/) (2026)
- [Infusion: Shaping Model Behavior by Editing Training Data via Influence Functions](https://mlsanity.com/paper/2602.09987/) (2026)
- [Beyond Owls: Subliminal Learning Can Transfer Learned Capabilities and Backdoors](https://mlsanity.com/paper/2610.10657/) (2026)

## Abstract (arXiv)

As machine learning models are increasingly fine-tuned on synthetic data, there is a critical risk of subtle misalignments spreading through interconnected AI systems. This paper investigates subliminal corruption, which we define as undesirable traits are transmitted through semantically neutral data, bypassing standard safety checks. While this phenomenon has been identified, a quantitative understanding of its dynamics is missing. To address this gap, we present a systematic study of the scaling laws, thresholds, and mechanisms of subliminal corruption using a teacher-student setup with GPT-2. Our experiments reveal three key findings: (1) subliminal corruption causes behavioral crossover, degrading the model's overall alignment, not just the targeted trait; (2) alignment fails in a sharp phase transition at a critical threshold of poisoned data, rather than degrading gradually; and (3) interpretability analysis shows the corruption mechanism mimics the model's natural fine-tuning process, making it difficult to detect. These results demonstrate a critical vulnerability in AI systems that rely on synthetic data and highlight the need for new safety protocols that can account for latent threats.