Continual Learning: Three Approaches That Work
Continual learning without catastrophic forgetting: what fruit flies, LoRAs and differentiable plasticity show about models that learn after training.
All
AI 70
Geology 26
Benchmarking 24
Machine Learning 17
Evaluation 12
Agentic AI 10
Open Source 9
Reinforcement Learning 5
China 4
Complexity Theory 3
Reinforcement Learning
Recursive Self-Improvement Is Hard for Big Labs
Why RL breaks down for recursive self-improvement at scale, and the two paths around it after six years building continuously-learning models.
Survival is the Only Reward
An open-ended AI feedback loop: an LLM survives by holding disk space in Linux containers and retrains on what worked. Sliding-window retraining won.
Survival is the Only Reward: Sustainable Self-Training Through Environment-Mediated Selection
The arXiv paper on negative-space learning: self-evolving agents whose behaviours are selected by environmental viability, not by a reward function.
Generalising from Self-Produced Data: Model Training Beyond Human Constraints
Generalising from self-produced data: empirical validation, ungameable rewards and autonomous retraining as a way past the limits of human data.