At the Sun’s Edge, 122 Dark Lines Turn Into Unexpected Light
A balloon-borne solar telescope found 122 spectral lines turning from absorption to emission before the Sun’s visible edge.
Unexpected result
A Deterministic Search Finds Long Cycles—or Splits a Number
A number-theory algorithm turns small modular cycles into a deterministic search for a factor, using smooth numbers to keep the search fast.
A 3-by-3 Matrix Breaks a Longstanding Rank-Minimization Conjecture
A three-by-three matrix overturns a 2011 rank-minimization conjecture, showing two natural ways to measure error can disagree.
JWST's Distant Galaxies Refuse to Fit the Usual Cosmic Shapes
JWST images sharpened the view of early galaxies, but two reliable methods disagree on whether many are disks, spheroids, or still taking shape.
Why a Known Quantum Circuit Can Still Hide Its Ensemble’s Secrets
A theoretical result finds that quantum measurement outcomes can resist cloning, even when researchers know the circuit that created them.
Why Some Cubic Systems Defy Every Stability-Checking Algorithm
A new proof shows why no algorithm can always decide whether sufficiently high-dimensional cubic equations settle at the origin.
Kepler’s 'Lonely' Planets May Be Members of Hidden Crowds
Kepler’s weak, lone planet signals may not mark empty systems: detection limits can make crowded planetary families look solitary.
The Planet-Forming Disk Where Methanol Appears Without Water
ALMA detected methanol but no water-vapor signal in one planet-forming disk, hinting that planetary systems may inherit different ice chemistry.
A Tiny Plus Solves a Grid Game That Defied Easy Answers
A simple plus-shaped tactic solves a long-standing grid game across every board size, with computer checks formally certified in Lean.
A Fading Word Trail Reveals an Unexpected Shift in Astronomy
A census of astronomy preprints finds a fading language trace as model-assisted writing grows—and explains what word counts can’t prove.
The AI Could Steer Its Reasoning—Once It Told Itself To
New experiments reveal that AI reasoning traces are easier to steer than earlier tests suggested—and that this control can complicate safety monitoring.
Four AI Teammates Beat Bigger Systems at RNA Sequence Design
A cooperative AI team divided RNA sequence design among four agents and beat established methods in computer simulations—with important limits.
The Cosmic Disk That Changed Its Stability When the Frame Shifted
A theoretical disk model flips from unstable to stable when its frame changes, revealing how simulation boundaries can imitate cosmic structure.
Why Repeating a Quantum Test Can Make Perfect Cheating Possible
A mathematical surprise shows how repeated quantum tests can turn an impossible one-shot conversion into certain success.
A Young Star’s Debris Disk Is Tilted, but the Culprit Is Missing
JWST reveals a 27-degree minimum tilt between a young star and its debris disk, while a giant-planet explanation remains elusive.
A Graph-Theory Surprise Shrinks the Search for Rare Geometric Cases
A graph-theory result links sparse shared-neighbor patterns to smaller epsilon-nets, revealing a surprising path around a logarithmic cost.
Why a Quantum Atom Chain Keeps Rebuilding a Crystal After a Quench
A laser-controlled atom chain repeatedly rebuilt its crystal pattern, revealing rare quantum states that resist the usual rush toward thermalization.
How an Invisible Intruder Can Split a River of Stars in Two
A dark subhalo may split a stellar stream into parallel strands, giving astronomers a new way to search for invisible matter.
A Random Walk’s Valleys Ignore How Far Each Step Goes
A random walk’s valley count can ignore jump lengths—but its stopping rule can transform the statistics entirely.
How Interference Makes a Photon’s Local Orbit Run Backward
A heralded single-photon experiment reveals how interference can reverse light’s local orbital flow in dim regions of a beam.