6 stories, newest first · one card per story
RAMA integrates episodic memory into multi-agent DRL to cut mean end-to-end latency by 13.4% and 99th-percentile latency by 66.7% in a 140-satellite Kepler constellation.
PRISM reports high AUROCs for emergency diagnosis, deterioration, ICU admission, and mortality.
SEER improves event-conditioned time-series forecasts with self-evolving retrieval and causal memory.
RADAR uses retrieval-augmented diffusion to learn market states for stochastic discount factor portfolio optimization.
VeriSpeak tests whether audio language models can verify 3,879 spoken claims, with retrieval plus reasoning reaching 86.1%.
A distillation framework compresses LLM reasoning into a 15.5M-parameter trade-up recommendation model reaching AUC 0.941 with product-type test-time training.
Today 1.6K readers · 2.4K views · 775 reads · all time 189K views