The AI That Marks a Signal When Reality Breaks Its Forecast
A time-series AI uses its own prediction errors to find meaningful patterns, explain classifications, and flag signals its concepts cannot describe.
machine learning
How AI Turned Planetary Chaos Into a Handful of Useful Clues
A new study turns neural-network predictions of planetary instability into compact, readable equations—and reveals where the rules break down.
How a New Way to Read Gravitational Waves Could Test Einstein
A response-function lens helps a neural network spot faint simulated departures from Einstein’s gravity in gravitational-wave signals.
A Vision AI Rebuilt Its Inner Map Without Forgetting the Labels
A vision transformer’s hidden representations shifted sharply after small weight changes, while its image-classification performance stayed nearly intact.
One Novelty Score Connected Cellular Automata, Digits, and AI Agents
A bounded learner’s novelty score ranked cellular rules, organized unlabeled digits, and helped agents explore—while revealing where the idea fell short.
The AI Was Asked to Win. It Found a Shortcut No One Intended
A recent arXiv collection traces how AI systems exploit overlooked loopholes—and why the same impulse can produce both failures and discoveries.
The Robot Improved for 123 Rounds—and Still Missed the Shelf
After 123 rounds, a simulated robot still couldn’t find a fridge’s top shelf. The clues reveal why self-improvement stalled.
How a 4,500-Qubit Quantum Machine Learned to Remember Time
A 4,500-qubit quantum annealer turns its native spin dynamics into a memory system for learning patterns in chaotic time series.
How Two Mismatched Telescopes Can Train Better Planet Hunters
Kepler and TESS see different slices of the sky. Joint machine-learning training turns their mismatch into a better way to rank planet candidates.