collaborators

5 papers

cs.LG2026

Discovery by Dreaming: Cross-Domain Recombination in Artificial Memory

Oliver Zahn, James Evans, David Eagleman

Dreams splice together people, places, and times that never met. Neuroscience suggests this recombination is not noise, but a function driving insight and creative discovery. This…

cs.AI2026

Facts as First Class Objects: Knowledge Objects for Persistent LLM Memory

Oliver Zahn, Simran Chana

Large language models increasingly serve as persistent knowledge workers, with in-context memory - facts stored in the prompt - as the default strategy. We benchmark in-context mem…

cs.AI2026

Selective Memory for Artificial Intelligence: Write-Time Gating with Hierarchical Archiving

Oliver Zahn, Simran Chana

Retrieval-augmented generation stores all content indiscriminately, degrading accuracy as noise accumulates. Parametric approaches compress knowledge into weights, precluding selec…

cs.LG2026

ANML: Attribution-Native Machine Learning with Guaranteed Robustness

Oliver Zahn, Matt Beton, Simran Chana

Frontier AI systems increasingly train on specialized expert data, from clinical records to proprietary research to curated datasets, yet current training pipelines treat all sampl…

q-bio.NC2026

Attention Is Not Retention: The Orthogonality Constraint in Infinite-Context Architectures

Oliver Zahn, Matt Beton, Simran Chana

Biological memory solves a problem that eludes current AI: storing specific episodic facts without corrupting general semantic knowledge. Complementary Learning Systems theory expl…