How these papers were chosen among the week's 2240 papers

All 2240 papers of the week whose main category is in this field were read, title and start of the abstract, and 92 were kept. They were then ordered by what can be counted: how many research profiles of our readers they come close to, whether a rising topic of the field includes them, and how many categories they were posted in. The abstracts of the first 40 were read in full to choose these, no more than two from one category and one per team.

The 1 number of the text was checked against the abstracts. 24 statements on what the papers found were looked for word for word in their abstracts.

Sources: arXiv.

An extremal black hole, reached through vacuum #

Can a black hole evolve into an extremal state, where its rotation reaches the limiting value allowed by its mass? A new preprint constructs a counterexample in vacuum to the third law of black-hole thermodynamics. The authors build initial conditions whose subsequent spacetime contains an exactly Schwarzschild apparent horizon and, later, an exactly extremal Kerr event horizon. The distinction between apparent and event horizons matters: the construction connects different kinds of horizon, rather than simply tracking a single locally defined surface.

The mathematical engine is a gluing theorem for Einstein’s vacuum equations. Rotating pulses of gravitational radiation raise the angular momentum, while a long intervening region suppresses unwanted incoming data. A final matching argument handles the extremal horizon exactly, including curvature information transverse to it. The result shows that vacuum general relativity permits a route that a familiar thermodynamic prohibition would appear to exclude; it is an existence proof, not a prediction that astrophysical black holes ordinarily follow this route.

The geometry of enclosing space #

The generalized Cartan–Hadamard conjecture concerns how efficiently a boundary can enclose volume in spaces with an upper curvature bound. This preprint reports a proof, together with rigidity: the equality case is constrained, rather than merely achieving the same numerical bound by accident. The authors also extend the result to Alexandrov spaces with curvature bounded on both sides, a setting that accommodates less smooth geometry.

The work brings several consequences under the same roof. Regions that minimize boundary size for their volume must be bounded, and hypersurfaces satisfy a sharp inequality of Michael–Simon type, linking geometric size to curvature. Such inequalities provide tools for studying surfaces and geometric differential equations. The abstract announces these results but does not describe the proof mechanism, so its central technical innovation cannot be assessed from the summary alone.

A speed limit for finding roots #

Root-finding algorithms repeatedly query a function or its derivatives to home in on a zero. The Kung–Traub conjecture asks how fast that convergence can be when each iteration has a fixed evaluation budget and carries no memory between iterations. A new preprint proves the proposed upper bound, even when an algorithm chooses its queries adaptively and processes the answers without restrictions on its rules.

The proof turns this broad algorithmic freedom into an adversarial problem. Polynomial Wronskians—determinants built from functions and derivatives—supply a lower bound at a fixed scale. A diagonal construction then produces an entire function that defeats any claimed higher order of convergence. The authors also report formal verification in Lean. Beyond numerical analysis, this is a result about information: clever processing cannot extract more local accuracy than the permitted queries support.

A community can fail despite reassuring interaction signs #

In mathematical ecology, qualitative permanence asks whether the signs of interactions can guarantee that species persist, rather than drift toward extinction. Earlier work identified necessary restrictions on feedback cycles in Lotka–Volterra interaction matrices and conjectured that those restrictions were sufficient. This preprint supplies a counterexample: an interaction pattern satisfies the proposed rules, yet supports dynamics that draw an open set of initially interior populations toward the boundary, where species disappear.

The mechanism is an attracting heteroclinic cycle, a sequence of connections between boundary equilibria. The example combines a directed feedback loop with a predator feeding on multiple prey. Most of its verification is elementary; a more difficult connection requires a lemma based on invasion graphs. The broader lesson is that a network’s positive and negative arrows can conceal dynamical routes to extinction, even when its cycle signs pass a plausible structural test.

Layering does not exhaust composite design #

Repeatedly layering constituent materials is a powerful way to design a composite, but can increasingly elaborate layering reproduce every attainable effective response? This preprint answers no for certain planar composites governed by linear systems with many potential components. The authors construct constituent properties for which every hierarchical laminate obeys an equality between effective energies, then build periodic microstructures that violate it.

The construction begins with locally layered material whose proportions vary smoothly, before replacing it with fine stripes of the original constituents. It also separates the quasiconvex and rank-one convex envelopes of a particular energy—mathematical objects used to describe relaxation and the onset of fine structure. The counterexample and the strict separation are formally verified in Lean 4. For materials design, the message is not that layering is ineffective, but that restricting a search to layered architectures can miss genuinely different responses.

A familiar law survives dependent coordinates #

The largest eigenvalue of a sample covariance matrix is important in statistics because it can signal a dominant direction of variation—or an unusually large fluctuation. A new preprint establishes the Tracy–Widom law for this eigenvalue under a concentration assumption on quadratic forms. The sampled vectors remain independent of one another, but coordinates within each vector may be dependent.

That distinction broadens the reach of a distribution already central to random matrix theory. The proof establishes edge universality: after suitable centering and scaling, the extreme eigenvalue follows the same limiting law across the allowed models. In particular, the result removes an unconditionality assumption from existing work on log-concave ensembles and extends beyond that class. It strengthens the mathematical basis for treating some high-dimensional spectral fluctuations as universal without assuming independent measurements within each observation.


These preprints examine the boundary between a useful rule and a universal one. Some establish limits that no algorithm or geometric competitor can evade; others expose configurations that defeat an accepted conjecture or a convenient design strategy. The common contribution is precision about assumptions: which ones support a theorem, which ones can be removed, and which intuitive conclusions fail without further control.

Rising in this field

All the rising topics →

Was this letter useful to you?

Written by AI from the 2710 papers of the week in this field. Most are preprints, not yet peer reviewed: the text says what their authors report.

Get one letter by email every week with a free account, all of them with Premium. Create a free account