# Pretraining Curricula Enable Selective Fine-tuning

arXiv: https://arxiv.org/abs/2607.04846 · HTML: https://mlsanity.com/paper/2607.04846/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 50 of its 71 references that are in the map.
- First posted on arXiv 6 Jul 2026.

## Builds on (most cited first, 30 of 50)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [On the Spectral Bias of Neural Networks](https://mlsanity.com/paper/1806.08734/) (2018)
- [Transformer Feed-Forward Layers Are Key-Value Memories](https://mlsanity.com/paper/2012.14913/) (2020)
- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations](https://mlsanity.com/paper/1811.12359/) (2018)
- [Towards Automated Circuit Discovery for Mechanistic Interpretability](https://mlsanity.com/paper/2304.14997/) (2023)
- [What Can Transformers Learn In-Context? A Case Study of Simple Function Classes](https://mlsanity.com/paper/2208.01066/) (2022)
- [Safety Alignment Should Be Made More Than Just a Few Tokens Deep](https://mlsanity.com/paper/2406.05946/) (2024)
- [Data Distributional Properties Drive Emergent In-Context Learning in Transformers](https://mlsanity.com/paper/2205.05055/) (2022)
- [Physics of Language Models: Part 3.1, Knowledge Storage and Extraction](https://mlsanity.com/paper/2309.14316/) (2023)
- [Curriculum Learning: A Survey](https://mlsanity.com/paper/2101.10382/) (2021)
- [How to use and interpret activation patching](https://mlsanity.com/paper/2404.15255/) (2024)
- [SGD on Neural Networks Learns Functions of Increasing Complexity](https://mlsanity.com/paper/1905.11604/) (2019)
- [How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model](https://mlsanity.com/paper/2305.00586/) (2023)
- [A mathematical theory of semantic development in deep neural networks](https://mlsanity.com/paper/1810.10531/) (2018)
- [What In-Context Learning "Learns" In-Context: Disentangling Task Recognition and Task Learning](https://mlsanity.com/paper/2305.09731/) (2023)
- [Does Circuit Analysis Interpretability Scale? Evidence from Multiple Choice Capabilities in Chinchilla](https://mlsanity.com/paper/2307.09458/) (2023)
- [The Quantization Model of Neural Scaling](https://mlsanity.com/paper/2303.13506/) (2023)
- [Skill-it! A Data-Driven Skills Framework for Understanding and Training Language Models](https://mlsanity.com/paper/2307.14430/) (2023)
- [The mechanistic basis of data dependence and abrupt learning in an in-context classification task](https://mlsanity.com/paper/2312.03002/) (2023)
- [Circuit Component Reuse Across Tasks in Transformer Language Models](https://mlsanity.com/paper/2310.08744/) (2023)
- [When Do Curricula Work?](https://mlsanity.com/paper/2012.03107/) (2020)
- [What needs to go right for an induction head? A mechanistic study of in-context learning circuits and their formation](https://mlsanity.com/paper/2404.07129/) (2024)
- [Findings of the BabyLM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corpora](https://mlsanity.com/paper/2504.08165/) (2025)

## Abstract (arXiv)

Transformers follow implicit curricula whereby some tasks are learned before others. However, how explicit pretraining curricula influence learning, generalization, and the selectivity of fine-tuning is unclear. This is important for AI safety, where fine-tuning is used to selectively suppress misaligned behaviors. Here, we compare curricula that pretrain tasks in a balanced (sampled uniformly) or an imbalanced (one task early, the other late) fashion. We show that imbalanced learning of two conflicting copy tasks promotes in-context learning and improves the selectivity of refusal fine-tuning. Ablations and activation patching show that this occurs because imbalanced pretraining encourages tasks to be disentangled in separable neural circuits, whereas balanced training routes both tasks through a common pathway. We extend these findings to a synthetic language learning task involving rule-consistent and rule-violating data, where imbalanced curricula similarly lead to more localized, less entangled rule representations, resulting in more robust rule-following behavior. Together, these results suggest that imbalanced pretraining curricula may be an important tool for promoting disentangled representations, with direct consequences for the precision and reliability of safety fine-tuning.