2 papers
cs.AI2026
Evaluating Agentic Learning Harness Capabilities Without Labels via the Scaling Hypothesis
Aryan Luthra, Kshitij Jain, Siddharth Arya +2
Agentic "Continual Learning Harnesses", systems that pair an LLM with retrieval or memory to improve from feedback without retraining, have shown growing value in cybersecurity. Bu…
cond-mat.mtrl-sci2026
Simulations of dislocation dynamics on an atomic lattice: the effect of collision rules
Tom Hudson, Akaraphon Jantaraphum, Patrick van Meurs
The stochastic dynamics of dislocations on a one-dimensional periodic lattice domain are considered. Two models are studied: one without a collision rule, and one which annihilates…