Search papers, labs, and topics across Lattice.
This paper introduces Abstraction-Augmented Training (AAT), a novel loss-level modification that encourages continual learning models to capture latent relational structures shared across examples. AAT jointly optimizes over concrete instances and their abstract representations, introducing a memory-efficient inductive bias that stabilizes learning in online data streams without replay buffers. Experiments on relational and narrative datasets demonstrate that AAT achieves performance comparable to or exceeding strong experience replay baselines, while requiring zero additional memory.
Forget replay buffers: this method achieves state-of-the-art continual learning performance with zero additional memory by training on both concrete data and its abstract representations.
The real world is non-stationary and infinitely complex, requiring intelligent agents to learn continually without the prohibitive cost of retraining from scratch. While online continual learning offers a framework for this setting, learning new information often interferes with previously acquired knowledge, causes forgetting and degraded generalization. To address this, we propose Abstraction-Augmented Training (AAT), a loss-level modification encouraging models to capture the latent relational structure shared across examples. By jointly optimizing over concrete instances and their abstract representations, AAT introduces a memory-efficient inductive bias that stabilizes learning in strictly online data streams, eliminating the need for a replay buffer. To capture the multi-faceted nature of abstraction, we introduce and evaluate AAT on two benchmarks: a controlled relational dataset where abstraction is realized through entity masking, and a narrative dataset where abstraction is expressed through shared proverbs. Our results show that AAT achieves performance comparable to or exceeding strong experience replay (ER) baselines, despite requiring zero additional memory and only minimal changes to the training objective. This work highlights structural abstraction as a powerful, memory-free alternative to ER.
