# Active perception and disentangled representations allow continual, episodic zero and few-shot learning

arXiv: https://arxiv.org/abs/2602.19355 · HTML: https://mlsanity.com/paper/2602.19355/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 41 references that are in the map.
- First posted on arXiv 22 Feb 2026.

## Builds on (most cited first, 20 of 20)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks](https://mlsanity.com/paper/1703.03400/) (2017)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Continual Lifelong Learning with Neural Networks: A Review](https://mlsanity.com/paper/1802.07569/) (2018)
- [An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks](https://mlsanity.com/paper/1312.6211/) (2013)
- [DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning](https://mlsanity.com/paper/2411.04983/) (2024)
- [Towards Continual Reinforcement Learning: A Review and Perspectives](https://mlsanity.com/paper/2012.13490/) (2020)
- [Neural Episodic Control](https://mlsanity.com/paper/1703.01988/) (2017)
- [MemRL: Self-Evolving Agents via Runtime Reinforcement Learning on Episodic Memory](https://mlsanity.com/paper/2601.03192/) (2026)
- [Model-Free Episodic Control](https://mlsanity.com/paper/1606.04460/) (2016)
- [Revisiting Active Perception](https://mlsanity.com/paper/1603.02729/) (2016)
- [Learning to Remember Rare Events](https://mlsanity.com/paper/1703.03129/) (2017)
- [Human-inspired Episodic Memory for Infinite Context LLMs](https://mlsanity.com/paper/2407.09450/) (2024)
- [One-shot Learning with Memory-Augmented Neural Networks](https://mlsanity.com/paper/1605.06065/) (2016)
- [Streaming Deep Reinforcement Learning Finally Works](https://mlsanity.com/paper/2410.14606/) (2024)
- [Questioning Representational Optimism in Deep Learning: The Fractured Entangled Representation Hypothesis](https://mlsanity.com/paper/2505.11581/) (2025)
- [Defining Benchmarks for Continual Few-Shot Learning](https://mlsanity.com/paper/2004.11967/) (2020)
- [The Kanerva Machine: A Generative Distributed Memory](https://mlsanity.com/paper/1804.01756/) (2018)
- [Long Distance Relationships without Time Travel: Boosting the Performance of a Sparse Predictive Autoencoder in Sequence Modeling](https://mlsanity.com/paper/1912.01116/) (2019)
- [Learning distant cause and effect using only local and immediate credit assignment](https://mlsanity.com/paper/1905.11589/) (2019)

## Abstract (arXiv)

Generalization is often regarded as an essential property of machine learning systems. However, perhaps not every component of a system needs to generalize. Training models for generalization typically produces entangled representations at the boundaries of entities or classes, which can lead to destructive interference when rapid, high-magnitude updates are required for continual or few-shot learning. Techniques for fast learning with non-interfering representations exist, but they generally fail to generalize. Here, we describe a Complementary Learning System (CLS) in which the fast learner entirely foregoes generalization in exchange for continual zero-shot and few-shot learning. Unlike most CLS approaches, which use episodic memory primarily for replay and consolidation, our fast, disentangled learner operates as a parallel reasoning system. The fast learner can overcome observation variability and uncertainty by leveraging a conventional slow, statistical learner within an active perception system: A contextual bias provided by the fast learner induces the slow learner to encode novel stimuli in familiar, generalized terms, enabling zero-shot and few-shot learning. This architecture demonstrates that fast, context-driven reasoning can coexist with slow, structured generalization, providing a pathway for robust continual learning.