Papers

The additivity of minimum output entropies is a central problem in quantum information theory. Nonadditivity is known for every Rényi order $p>1$, at the von Neumann point $p=1$, and for sufficiently small positive $p$, while much of the interval $0<p<1$ has remained open. In this work, we prove that for every $p>0$ there exist finite-dimensional quantum channels whose minimum output $p$-Rényi entropies violate additivity. Our proof combines two constructions: random projection-induced channels yield nonadditivity for $p>3/4$, and antisymmetric postprocessing extends the violation to all positive Rényi orders. Our estimates also improve the output-dimension bound obtained by Belinschi, Collins, and Nechita for additivity violation of minimum output von Neumann entropy.

Unadjusted samplers such as unadjusted Hamiltonian Monte Carlo and underdamped Langevin are well-known to be biased. Metropolis–Hastings adjustment has been conventionally incorporated into Hamiltonian Monte Carlo to eliminate the bias. However, this adjustment can significantly increase the iteration complexity due to the small step size required for reasonable Metropolis acceptance rates. In this work, we extend the delocalization of bias phenomenon, previously established for the overdamped Langevin algorithm, to these two unadjusted algorithms. We show that to control the $W_2$ bias of any $K$-dimensional marginal of a high-dimensional distribution, $O(\sqrt{K})$ integration steps suffice up to $\log d$ terms, assuming either weak or sparse interactions among variables. The discrete-time integrators here introduce technical difficulties beyond those of the overdamped setting, which we address through a broadly applicable matrix-polynomial framework that characterizes their propagators. Our result for the underdamped Langevin algorithm is valid for all large friction parameters, implying that the Leimkuhler-Matthews integrator for the overdamped Langevin dynamics also exhibits delocalization of bias.

We generalize work of Wei-Wu on Viehweg hyperbolicity for families of stable pairs over smooth projective varieties to the case of smooth Deligne-Mumford stacks. The results of Wei-Wu build on results of Popa-Schnell and Viehweg-Zuo, utilize a result of Campana-Păun, and extend results of Kebekus-Kovács and Patakfalvi. We apply our results to smooth proper moduli stacks parameterizing generically klt relatively simple normal crossings KSBA stable pairs, showing that such moduli stacks naturally have big log canonical bundles. We also consider implications for their coarse moduli spaces.

This paper introduces a dynamic portfolio optimization framework for large institutional investors using Scientific Physics-Informed Reinforcement Learning (SciPhyRL). Formulated in continuous time over an extended state space that includes explicit cumulative costs, the approach leverages offline historical data to learn optimal, distribution-aware strategies. A core innovation reduces the optimization challenge to solving an HJB equation by projecting it onto observed trajectories as a pathwise Hamilton-Jacobi equation. This is solved directly from data using PINN in a single offline sweep, eliminating the need for traditional value or policy iteration. To make the method effective at practical short horizons, the control variable is recast from a continuous trading rate to a discrete target holding. This ensures signal-implied positions are reached immediately, while execution costs are evaluated against a microstructure-grounded quadratic price impact model. Evaluated on a $14$-asset ETF universe using an engineered oracle signal, the learned Gibbs policy yields substantial out-of-sample Sharpe ratio improvements over static and myopic baselines. The results demonstrate that the proposed framework successfully translates known signal quality into a robust, multi-period, and cost-aware allocation mechanism with strictly controlled volatility and turnover.

In this paper, we construct an efficient higher-order local time integration scheme for spatially discretized linear Friedrichs' systems. In particular, our interest is in problems where only a few of the mesh elements are small while the majority of the elements is much larger. The special combination of two methods like the leapfrog method on the coarse part of the mesh and the Crank-Nicolson method on the fine part as was done in Hochbruck, Sturm 2016 and Hochbruck, Köhler 2022 is not suitable for higher-order time integration. Therefore, we suggest to approximate the solution of the linear systems arising in each time step by a preconditioned Krylov subspace method, e.g., the quasi-minimal residual method by Freund and Nachtigal 1991. The techniques developed here for linear problems also carry over to nonlinear problems, where linear systems of the same type arise within a Newton-type iteration. Motivated by the analysis of locally implicit methods by Hochbruck and Sturm 2016, we show how to construct a preconditioner in such a way that the number of iterations required by the Krylov subspace method to achieve a certain accuracy is bounded independently of the diameter of the small mesh elements. We prove this behavior by using Faber polynomials and complex approximation theory. The cost to apply the preconditioner consists of the solution of a small linear system, whose dimension corresponds to the degrees of freedom within the fine part of the mesh (and its next coarse neighbors). If this dimension is small compared to the size of the full mesh, the preconditioner is very efficient. We conclude by verifying our theoretical results with numerical experiments for the fourth-order Gauss-Legendre Runge–Kutta method.

We study the optimal constants in higher-order Hardy inequalities on the lattice $\mathbb{Z}^d$. For each fixed $\ell \in \mathbb{N}$, we prove that the optimal constant $\mathcal{C}_\text{opt}^\ell(d)$ in $$ \sum_{n \in \mathbb{Z}^d} |\Delta^{\ell/2}u(n)|^2 \geq \mathcal{C}_\text{opt}^\ell(d)\sum_{n \in \mathbb{Z}^d} \frac{|u(n)|^2}{|n|^{2\ell}}. $$ satisfies $$ \lim_{d\rightarrow\infty}\frac{\mathcal{C}_\text{opt}^\ell(d)}{d^\ell} =2^\ell. $$ The proof is based on a Fourier reduction to a family of singular Hardy inequalities on the flat torus, involving the weight \[ \omega(x)^{-2\ell}, \qquad \omega(x)^2=\sum_{j=1}^d\sin^2\left(\frac{x_j}{2}\right), \] and zero average condition on admissible functions. We establish these torus inequalities by combining a ground state representation formula with a weighted integrated Bochner identity in an iterative scheme. The method yields explicit constants, defined recursively in the order $\ell$, and requires only the classical unweighted Poincaré inequality on the torus. The appearance of the limiting constant $2^\ell$ is particularly striking, as it suggests that, in the high dimensional regime, the optimizers are localized near the unit sphere $\{n\in\mathbb{Z}^d:|n|=1\}$ in $\mathbb{Z}^d$.

Inference for infectious disease transmission on dynamic contact networks is complicated by latent infection times, partially observed network evolution, measurement error in contact data, and infection originating from outside the observed population. Existing likelihood-based approaches typically address these challenges separately and often rely on restrictive assumptions such as fully observed networks, closed populations, or symptom onset as a surrogate for infection time. We develop a unified complete-data likelihood framework for epidemic processes evolving on partially observed dynamic networks. The proposed formulation represents disease progression, network evolution, and observation mechanisms as interacting continuous-time stochastic processes within a common probabilistic framework. Specifically, we couple a susceptible-exposed-infectious-removed (SEIR) epidemic process with a status-dependent dynamic contact network and explicit observation models for symptoms and contacts. The resulting framework accommodates latent incubation periods, intermittent network observation, contact measurement error, and external infection pressure while preserving a coherent likelihood structure. Our principal contribution is the derivation of a complete-data event-history likelihood for the joint epidemic-network process under partial observation. The likelihood provides a rigorous foundation for likelihood-based and Bayesian inference through data augmentation, clarifies how information from disease progression and contact dynamics jointly determines parameter estimability, and reveals a broad class of existing epidemic network models as special cases. More generally, the framework contributes to statistical inference for partially observed interacting stochastic systems on evolving networks and establishes a foundation for uncertainty-aware analysis of complex transmission processes.

Coulomb branches of 3-dimensional $N=4$ gauge theories, as defined by Braverman–Finkelberg–Nakajima, form a large and interesting class of symplectic singularities. In this paper, we give a geometric description of these varieties in terms of blowups of toric compactifications, uniformly across the rational, $K$-theoretic, and elliptic settings, which in physics are often associated with 3-, 4-, and 5-dimensional versions of these theories. We use the resulting local description to classify the symplectic leaves of Coulomb branches and describe their transverse slices in terms of the classification of zero-dimensional leaves of certain smaller Coulomb branches. We also discuss related topics, including removing an unnecessary hypothesis from a theorem of Gannon and the author on functoriality for Coulomb branches, and clarifying the relationship with transverse Hilbert schemes by refining the approach of Bielawski and Foscolo.

In this paper, we study the optimization landscape induced by the true loss for shallow polynomial neural networks (PNNs) with $\mathfrak{h} \in \mathbb{N}$ neurons on the hidden layer, one-dimensional input and output layers, and a monomial activation of degree $d \in \mathbb{N}$, trained against a non-constant affine linear target function. Our first main result provides for arbitrary activation degree $d$ a sharp existence/non-existence criterion for global minimizers with necessary structural conditions. We show that the infimum of the loss is always zero and achievable with at least $d$ active and visible hidden neurons – that is, hidden neurons with non-zero inner and outer weights – with pairwise distinct pivots. In contrast, if $\mathfrak{h} < d$, then the infimum cannot be attained and any minimizing sequence of parameters necessarily diverges to infinity. In the second main result, we provide a complete classification of all critical points of the loss function for the cubic activation. We show that the loss landscape admits no local maximizers, critical points cannot have exactly two distinct pivots, global minimizers require at least three distinct pivots, critical points with no active hidden neurons correspond to saddle points only, and consequently, non-global local minimizers and non-trivial saddle points arise only in networks where all pivots coincide. Moreover, non-global local minimizers require all hidden neurons to be active and visible with exactly one hidden neuron having a slope sign matching that of the target function. Our second main result also guarantees that each hidden neuron of a critical point that is not a global minimizer has either input-dependent or zero contribution, but has no nonzero input-independent contribution, to its corresponding realization function.

Many computational models arising in science and engineering exhibit a multiscale structure that makes the assembly or direct solution of the global problem computationally prohibitive. Domain Decomposition (DD) methods overcome this limitation by replacing the global problem with a sequence of coupled local problems, whose iterative solution reconstructs the global response. This work introduces a method in the family of Domain Decomposition Reduced Order Models (DD-ROMs), based on the observation that DD naturally localizes not only the solution operator but also its geometric and parametric dependence. The central idea is that DD transforms a globally intractable solution map into a family of locally representable operators learnable from affordable local data after identification with a common reference configuration, a concept that we formalize through the notion of local representability. Non-intrusive neural surrogates are then trained to approximate the fine-scale local operations and embedded into the iterative solver. The training algorithm is based on a cascaded strategy designed to match the distributions encountered by the deployed surrogate iteration. We interpret the resulting DD method as a perturbed fixed-point iteration and establish that the global error remains bounded by the surrogate approximation error. The framework is instantiated for mixed-dimensional elliptic problems coupling three-dimensional bulk domains with embedded one-dimensional inclusions, using a two-level non-overlapping Robin-Robin method. Numerical experiments show that the resulting DD-ROM is stable, achieves accurate approximation on unseen microscale geometries and features good scalability properties with respect to the number of subdomains, scaling to large size global problems while avoiding fine-scale operator assembly and local high-fidelity solvers in the online stage.

We study locally flat, compact, oriented, genus g, surfaces with boundary a fixed knot K, properly embedded in simply-connected 4-manifolds with boundary the 3-sphere, and whose complements have trivial fundamental group. We show that if two such surfaces are homologous then they are topologically ambiently isotopic rel. boundary.

We construct a new family of Calderbank-Shor-Steane (CSS) codes using the generator and parity-check matrices of Low-Density Generator Matrix (LDGM) codes, with row operations applied to both matrices in order to achieve the desired quantum rate. Decoding is performed in an iterative manner, by applying message passing over the associated graph, and discrete Density Evolution (DDE) is used to optimize performance in the depolarizing channel. The proposed construction offers high flexibility and easiness in the design, producing quantum codes that possess excellent error correction capabilities. By properly designing the structure of the code, we are able to control and bound the weight of the stabilizer generators to a small value, which results in codes particularly well suited for fault-tolerant quantum computation. At the same time, these codes achieve very good performance in terms of error correction capability.

We study the Grothendieck ring of finite-dimensional non-degenerate evolution algebras over a field $\mathbb{K}$, with addition induced by direct sum and multiplication induced by tensor product. Although its underlying abelian group is freely generated by indecomposable isomorphism classes, the ring structure is much smaller: tensor products create systematic non-cancellation phenomena and many zero-divisors. The key invariant is the balance of the directed graph associated with a natural basis. We prove that, for non-degenerate evolution algebras, this balance is independent of the chosen natural basis. We then show that balance controls cancellation after tensoring: indecomposable factors of balance $1$ cancel, whereas factors of larger balance produce zero-divisor relations under mild hypotheses. Finally, we analyze the subring generated by cyclic evolution algebras and prove the predicted zero-divisor criterion for all finite direct sums of cyclic algebras.

We investigate the ultradifferentiable regularity of solutions and vectors of a class of hypoelliptic differential operators first introduced by Hörmander. Our results extend results of Hashimoto–Morimoto–Matsuzawa and Bolley–Camus–Metivier.

The entanglement-assisted capacity of a quantum channel admits an additive single-letter characterization, implying that joint encodings across channel uses cannot increase the ultimate communication rate. Here, we show that this additive picture does not extend to communication reliability. Specifically, we prove that the Petz-Rényi channel information can be strictly superadditive for every $\alpha\in[1/2,1)$, yielding a genuine multi-copy enhancement of the entanglement-assisted random-coding error exponent, even though the entanglement-assisted capacity remains additive. We establish this phenomenon analytically already for measurement channels, which are entanglement-breaking and have additive unassisted capacity. Remarkably, this strict superadditivity is witnessed by a separable, classically correlated two-copy channel-input marginal, demonstrating that no entanglement between the transmitted systems is required. Our results show that, although correlations across channel uses cannot increase the ultimate rate of entanglement-assisted communication, they can enhance its reliability.

A modular continuous aperture array (CAPA)-based multi-user communication system is investigated, where only a portion of the aperture, namely sub-CAPAs, is activated to serve users. The signal model for the proposed modular CAPA is first introduced. Based on this model, a spectral efficiency (SE) maximization problem is formulated to jointly optimize the sub-CAPA activation and beamforming, subject to constraints on the limited number of active sub-CAPAs and the total transmit power. To address the resulting mixed-integer optimization problem, a branch-and-bound (B&B)-based algorithm is first proposed for optimal sub-CAPA activation and beamforming design. After that, the spatial bandwidth of the modular CAPA under partial activation is analyzed. The analysis reveals that a modular CAPA with partial sub-CAPAs activated could achieve a maximum spatial bandwidth comparable to that of a conventional CAPA. Motivated by this insight, a low-complexity spatial bandwidth-aware sub-CAPA activation scheme is further proposed. Finally, numerical results demonstrate that i) modular CAPA architectures with partial activation can consistently achieve greater performance gains than adjacent CAPA activations; ii) the proposed B&B scheme outperforms all benchmark schemes in terms of SE; and iii) the proposed spatial bandwidth-aware scheme provides an attractive performance-complexity tradeoff compared with the proposed B&B-based algorithm.

We prove a Weyl law for the number of quasinormal modes (QNM) of a Schwarzschild black hole contained in a sector below the real axis. This requires introducing a new pseudodifferential operator calculus tailored to the study of semiclassical spectral problems near threshold energies. Elliptic theory in this calculus can be combined with the method of complex scaling to give uniform resolvent estimates near zero energy for operators that behave at infinity like a semiclassical Schrödinger operator with a repulsive inverse-square potential. Applied to the Regge-Wheeler potential, our methods imply the absence of high angular momentum QNM from a disc whose radius grows linearly with the angular momentum. Together with the asymptotic description of Schwarzschild QNM recently obtained by Hitrik and Zworski, this shows that the number of QNM contained in a small sector below the real axis and with modulus bounded by $\lambda$ grows as $C\lambda^3$. We also study the effect of cutting off the Schwarzschild resolvent away from the event horizon and show that such a cutoff does not lead to any pole cancellations.

To numerically integrate a function, one may resort to Quasi-Monte Carlo estimators, that average integrand values at pseudo-random well-distributed uniform sampling locations. Better uniformity improves the worst-case integration-error bound. A standard measure of uniformity is given by an integer $t$ value, where $t=0$ yields the best uniformity. Producing sequences of samples with bounded $t$ values can be achieved with Sobol' recursive construction, that uses coefficients of irreducible polynomials. While $b$-dimensional sequences with $t=0$ can be obtained by taking $b$ polynomials of degree $1$ over the Galois Field $\mathrm{GF}(b)$, we show conditions that guarantee $t=0$ for specific higher degree polynomials. In particular, we relate the Sobol' construction to tensorized powers of Pascal matrices when the chosen polynomials only differ by a constant and exhibit simple conditions to guarantee $t=0$ in this case. We then focus on Artin-Schreier irreducible polynomials, in the form $p_i(x) = x^b - x + c_i$, where $i \in \{1, \dots, b-1\}$ and $b$ is prime, and we make explicit conditions that always guarantees $t=0$ in $b-1$ dimensions. Combining $b$-dimensional Sobol' of degree $1$ and our $(b-1)$-dimensional Artin-Schreier sequence of degree $b$, we provide a fast greedy procedure that optimizes the $(2b-1)$-dimensional combined $t$ value, while guaranteeing $t=0$ projection in subspaces.

Establishing the global-in-time smoothness of solutions to the 3D Navier-Stokes equations (NSEs) with large initial data remains a long-standing open problem. In this paper, we prove that the 3D NSEs driven by a suitably chosen transport noise – a physically motivated stochastic perturbation – admit global-in-time smooth solutions with high probability. This regularizing effect holds uniformly for initial data in arbitrarily large balls of subcritical function spaces with positive smoothness.

We prove a Bombieri-Vinogradov type theorem for exponential sums over products of $k$ primes. As an application, we show the lower bound $$\sup_{n\le x} \left|\sum_{m\le n} 1_{\Omega(m) = k} e(\alpha m)\right| \gg x^{1/6 - \varepsilon}$$ for $2\le k\le (2-\varepsilon)\log\log x$ and $\alpha\in\mathbb{R},$ where we noted $e(\beta) := e^{2i\pi\beta}.$

The main theme of this paper is to study symplectic (un-)knottedness of Lagrangian submanifolds which are not tori. Our construction of such Lagrangians is inspired by the generalized Polterovich Lagrangians of Oakley–Usher and relies on an adaptation of symplectic reduction methods going back to Chekanov–Schlenk and McDuff's probes to the case of non-abelian group actions. We compare Oakley–Usher Lagrangians to their (un-)knotted cousins to find that, depending on the situation, they are: (1) not diffeomorphic, (2) diffeomorphic, but not Lagrangian isotopic, (3) Lagrangian isotopic, but not Hamiltonian isotopic, (4) Hamiltonian isotopic. This relationship between them frequently changes when the ambient space is compactified by a symplectic cut. In the companion paper "Lagrangian tori in the three-dimensional quadric and mirror symmetry", we focus on monotone tori in the three-dimensional quadric and compute their enumerative invariants and study their relationship to mirror symmetry.

We investigate a second-order Modica-Mortola functional of the form \[ E_\varepsilon[u] := \int_\Omega \frac{1}{\varepsilon} W(x, \nabla u) + \varepsilon|\nabla^2 u|^2 \, dx, \] which models solid–solid phase transitions in heterogeneous media. We neglect the assumption of frame indifference in the elastic energy density $W$ while allowing for space-dependent, pointwise-compatible wells. Under suitable assumptions on $W$ and regularity conditions on the associated rank-one connection, we prove that $E_\varepsilon$ $\Gamma$-converges to a local interfacial energy defined for suitable laminate-type configurations.

We derive the long time asymptotic behavior of the Laplace transform of the Elephant Random Walk (ERW). The ERW is a generalization of the Simple Random Walk in which the past influences the future evolution of the process according to some memory parameter. Its transition kernel is neither space nor time homogeneous and its transition probability at the $n$th step can be written as a function of $n$ and the position of the walk at time $n - 1$. A recurrence relation for its Laplace transform allows one to find a generating function governed by a transport-type PDE whose solution can be written in terms of a Schwarz-Christoffel mapping, allowing a precise analytic control. This enables to apply singularity analysis methods to estimate the Laplace transforms.

Let $T_k$ be the minimum positive integer $t$ such that, for every positive integer $n$, every equinumerous $t$-coloring of $[tn]$ contains a rainbow $k$-term arithmetic progression. Jungić, Licht, Mahdian, Nešetřil and Radoičić conjectured that $T_k=\Theta(k^2)$, while Conlon, Fox and Sudakov proved that $T_k=O(k^2\log k)$. We prove the matching lower bound $T_k=\Omega(k^2\log k)$, and hence $T_k=\Theta(k^2\log k)$.

This paper develops a categorical framework – Learning in Infinitesimal Non-Compositional Sketches (LINCS) – as the repair of non-compositionality: failures of diagrams to factor through quotient sketches lifted to the tangent category setting. Machine learning problems are specified as sketches: graphs with commutativity conditions $\mathcal D$, limit cones $\mathcal L$, and colimit cocones $\mathcal K$, generalizing the usual scalarization of loss functions or vector space assumptions. Non-compositionality is defined purely as failure of a universal factorization problem, not as arithmetic error between the desired and actual predictions. Given a learning sketch $\mathbb S=(S,\mathcal D,\mathcal L,\mathcal K)$, whose underlying graph is $S$, and a model $D:J \rightarrow C$, the base defect is the obstruction to factorization $\mbox{Obs}(\mbox{Fact}_{\mathbb S}(D))$. The tangent lift applies the tangent functor $T$ to obtain $TD:J \rightarrow C$, and LINCS is defined as the obstruction $\mbox{Obs}(\mbox{Fact}_{\mathbb S}(TD))$ – asking whether infinitesimal perturbations preserve the compositionality constraints.The paper also introduces Tangent Learning Sketches, which are sketches equipped with Cockett-Cruttwell tangent structure. The paper defines the INC endofunctor, which iterates the tangent lift, producing a tower $D,TD,T^2D, \cdots$ of factorization problems. ML is thereby formulated as the search for a coalgebraic fixed point where successive tangent unfoldings stabilize ($\nu T_{\mbox{INC}}$). Using the Aczel–Mendler theorem, we prove existence of a final INC coalgebra whenever $T_{\mbox{INC}}$ admits a set-based class realization that creates its final carrier. A detailed experimental evaluation of LINCS is underway in a number of concrete ML settings, including deep learning, large language models, and reinforcement learning, and is described in companion papers.

We obtain a quantitative form of Schur's comparison theorem for curves with finite total curvature in CAT(k) spaces. This sharpens and extends the classical arm and bow lemmas in Euclidean space, as well as their Riemannian analogues. The proof is based on a comparison formula for curves in model planes, expressed in terms of curvature measures and a notion of moment arm borrowed from mechanics. Another ingredient is a refinement of Reshetnyak's theorem that controls the curvature of the majorizing curve.

We construct a smooth complex projective variety whose Albanese morphism is a homotopy fiber bundle but not a submersion. The same variety fibers smoothly over the circle, although every holomorphic one-form on it has a zero. A second construction yields smooth complex projective varieties $X$ such that the Aomoto complex of every nonzero holomorphic one-form on every connected finite étale cover of $X$ is exact, while $X$ admits no real closed one-form without zeros. The two constructions build, respectively, on a homology fiber bundle of Corrêa–Kollár that is not a homotopy fiber bundle and on a rational cohomology torus constructed by Debarre–Jiang–Lahoz. Consequently, we disprove Kotschick's conjecture, the remaining implication in the Bobadilla–Kollár conjecture, and a conjecture of the first-named author.

We construct strongly convergent finite-dimensional random matrix models for finite $q$-Gaussian family in the range $\vert q \vert < \sqrt{2}-1$. The construction has two stages. First, we show that normalized sums of graph-product semicirculars over an Erdos-Renyi graph satisfy the $q$-Toeplitz relations up to an operator norm error converging to zero in probability. Using ultraproduct methods, this yields complete strong convergence, uniformly over all matrix coefficient dimensions, for noncommutative polynomials with bounded degree. Second, we use a quantitative tensor-GUE for graph-product semicirculars to convert these operator models into finite-dimensional random matrices. For every fixed polynomial degree, the convergence is uniform over coefficient dimensions that may be larger than the dimension of the random matrices. As applications, in the above range of $q$, the C$^\ast$-algebra generated by a finite $q$-Gaussian family is MF, and the Brown-Douglas-Fillmore extension semigroup of the nontrivial C$^\ast$-algebra generated by a finite $q$-Gaussian family is not a group.

Let $(X,\mathcal E)$ and $(Y,\mathcal F)$ be uniformly locally finite coarse spaces. We prove that every $C^*$-algebra isomorphism $C_u^*(X,\mathcal E)\cong C_u^*(Y,\mathcal F)$ forces $(X,\mathcal E)$ and $(Y,\mathcal F)$ to be bijectively coarsely equivalent. This completely resolves the isomorphism rigidity problem for uniform Roe algebras over arbitrary uniformly locally finite coarse spaces.

Comonads on Set generalize both categories and topological spaces. Expanding upon Garner's work on ionads, we develop aspects of the theory of topological spaces for arbitrary comonads on arbitrary categories. Our approach is centered around density comonads, which provide an abstraction of subbases. We study subbases as well as bases in terms of density comonads, and we study continuous maps of comonads in terms of functors between coalgebra categories, with definitions that recover the usual notions for topological spaces. Whereas Ahman and Uustalu characterized categories as precisely the polynomial comonads on Set, we characterize topological spaces as precisely the density comonads of diagrams of subsets of a set, which are familiar as topological subbases. We show that every comonad on Set has an underlying topological space, and that this construction is a reflection with respect to continuous maps; similarly, every comonad on Set has an underlying small category, and this construction is a coreflection. We also show that the category of all comonads on Set with continuous maps is complete, and that its full subcategory of accessible comonads is cocomplete. Continuous maps and ordinary comonad morphisms form a double category, which, in the case of the polynomial comonads on Set, recovers the double category of functors and retrofunctors of Clarke and Di Meglio. We find topological intuition for these concepts in terms of "halos", an abstraction of infinitesimal neighborhoods of points, defined as formal limits of neighborhood systems. We include a long appendix of counterexamples, many applicable to general (co)monad theory rather than the particular concerns of this text.