SearcharxivSearch

arXiv subjects

Joshua Steier

Publications and source records attributed to Joshua Steier.

10 recordsLinked to original sources

Cofilling Shattering: A Syndrome-Support Hierarchy for Check Erasures

Let $A:\mathbb{F}_2^n\to\mathbb{F}_2^m$ be a binary linear map with fixed coordinate bases, let $C_A=\ker A$, and let $\lambda_A(y)$ be the minimum Hamming weight of a preimage of the syndrome $y$. We define $\operatorname{Shat}_{q,s}(A)$ as the least common check support of a $q$-dimensional syndrome subspace whose every nonzero element has coset-leader weight at least $s$. It therefore distinguishes release of $q$ independent syndromes from release of a subspace with no easy linear combination. Deleting check coordinates $F$ releases $\ker A_{\bar{F}}/\ker A$, canonically isomorphic to $(\operatorname{im} A)[F]$. Finiteness implies $R_q(C_A)\ge \mathsf{N}_2(q,s)$, where $\mathsf{N}_2(q,s)$ is the shortest length of a binary code of dimension $q$ and distance at least $s$; profile-Griesmer bounds independently control common check support. The hierarchy is coordinate-relabeling invariant but can change under a change of check basis. For the pair-repetition code $C_n=\{(x,x):x\in\mathbb{F}_2^n\}$, the standard realization $H_0=[I_n\ I_n]$ has $\operatorname{Shat}_{q,s}(H_0)=\mathsf{N}_2(q,s)$ whenever feasible. For every $q\ge 1$ and $s\ge 2$, with $n=\mathsf{N}_2(q,s)$, a row-equivalent realization of the same code has value $q$. For a simplicial coboundary map $A=\delta_k$, check erasure is top-face erasure and the released quotient is emergent cohomology. At $s=1$ the hierarchy reduces to generalized Hamming weights and is Tutte-determined; for $s\ge 2$, even identical labeled cut codes can have different values.

cs.IT

Ergodic Deviation-Robust Equilibrium under Mirror Descent Learning in Finite Games

We introduce Ergodic Deviation-Robust Equilibrium (EDRE), a dynamics-relative equilibrium concept for repeated finite games in which agents learn via entropic mirror descent (EMD). EDRE requires three properties to hold simultaneously for the same profile and learning run: (E1) the limit profile is an $\varepsilon$-Nash equilibrium at a product distribution; (E2) along the entire learning trajectory, every fixed coalition's cumulative aggregate (summed-unilateral) deviation gain is $\tilde{\mathcal{O}}(\sqrt{T})$ with high probability; and (E3) the limit profile is a fixed point of the EMD map, so that it is selected by the dynamics rather than merely certified as an equilibrium. We prove that the $\sqrt{T}$ deviation-regret rate is order-tight, establish existence in exact-potential games (via Nash's theorem, with a constructive proximal route under concavity) together with Lyapunov monotonicity of EMD (and pointwise convergence when the fixed-point set is a singleton), and extend the selection property to monotone polymatrix games through variational inequalities. Although a static EDRE coincides with an $\varepsilon$-Nash equilibrium, its content is dynamic: robust (positive-measure) selection under EMD excludes linearly unstable equilibria, so EDRE acts as a Nash equilibrium equipped with a dynamic certificate rather than a static refinement. On the complexity side, we show that computing EDRE is PPAD-hard in general polymatrix games and belongs to promise-PPAD for potential games. A worked $2\times 2$ coordination-game example illustrates all components of the framework. Additional results, including a bandit-feedback extension, a period-doubling route to Li-Yorke chaos for the two-strategy EMD map at large step size, a linear-program formulation for minimum-cost steering, and supporting simulations, appear in the appendices.

cs.GT

Adjacency Spectral Radius Under Laplacian Sparsification: Deterministic and Probabilistic Bounds

Spielman-Srivastava spectral sparsification preserves Laplacian quadratic forms to within (1 +/- epsilon), but does not directly control the adjacency spectral radius lambda_1, which governs the NIMFA epidemic threshold and arises in spectral clustering. We prove |lambda_1(A_H) - lambda_1(A_G)| <= epsilon(2 Delta - lambda_1) deterministically, with a sharp epsilon*lambda_1 bound for reweighting sparsifiers via Perron-Frobenius monotonicity. Under effective-resistance sampling, Matrix Bernstein gives O(epsilon Delta / sqrt(c)) with high probability. Combining eigenvector delocalization with resolvent perturbation theory, we establish that for graphs with delocalized Perron eigenvectors and spectral gap = Omega(Delta), the distortion is O(epsilon Delta sqrt(log n) / sqrt(n)) + O(epsilon^2 Delta^2 / delta_gap), with corollaries for Erdos-Renyi graphs, regular expanders, and stochastic block models. Lower bounds establish tightness for regular graphs.

math.CO

Information Routing in Atomistic Foundation Models: How Task Alignment and Equivariance Shape Linear Disentanglement

What determines whether a molecular property prediction model organizes its representations so that geometric and compositional information can be cleanly separated? We introduce Compositional Probe Decomposition (CPD), which linearly projects out composition signal and measures how much geometric information remains accessible to a Ridge probe. We validate CPD with four independent checks, including a structural isomer benchmark where compositional projections score at chance while geometric residuals reach 94.6\% pairwise classification accuracy. Across ten models from five architectural families on QM9, we find a \emph{linear accessibility gradient}: models differ by $6.6\times$ in geometric information accessible after composition removal ($R^2_{\mathrm{geom}}$ from 0.081 to 0.533 for HOMO-LUMO gap). Three factors explain this gradient. Task alignment dominates: models trained on HOMO-LUMO gap ($R^2_{\mathrm{geom}}$ 0.44--0.53) outscore energy-trained models by $\sim$0.25 $R^2$ regardless of architecture. Within-architecture ablations on two independent architectures confirm this: PaiNN drops from 0.53 to 0.31 when retrained on energy, and MACE drops from 0.44 to 0.08. Data diversity partially compensates for misaligned objectives, with MACE pretrained on MPTraj (0.36) outperforming QM9-only energy models. Inside MACE's representations, information routes by symmetry type: $L{=}1$ (vector) channels preferentially encode dipole moment ($R^2 = 0.59$ vs.\ 0.38 in $L{=}0$), while $L{=}0$ (scalar) channels encode HOMO-LUMO gap ($R^2 = 0.76$ vs.\ 0.34 in $L{=}1$). This pattern is absent in ViSNet. We also show that nonlinear probes produce misleading results on residualized representations, recovering $R^2 = 0.68$--$0.95$ on a purely compositional target, and recommend linear probes for this setting.

cs.LG

When Does Margin Clamping Affect Training Variance? Dataset-Dependent Effects in Contrastive Forward-Forward Learning

Contrastive Forward-Forward (CFF) learning trains Vision Transformers layer by layer against supervised contrastive objectives. CFF training can be sensitive to random seed, but the sources of this instability are poorly understood. We focus on one implementation detail: the positive-pair margin in the contrastive loss is applied through saturating similarity clamping, $\min(s + m,\, 1)$. We prove that an alternative formulation, subtracting the margin after the log-probability, is gradient-neutral under the mean-over-positives reduction. On CIFAR-10 ($2 \times 2$ factorial, $n{=}7$ seeds per cell), clamping produces $5.90\times$ higher pooled test-accuracy variance ($p{=}0.003$) with no difference in mean accuracy. Analyses of clamp activation rates, layerwise gradient norms, and a reduced-margin probe point to saturation-driven gradient truncation at early layers. The effect does not transfer cleanly to other datasets: on CIFAR-100, SVHN, and Fashion-MNIST, clamping produces equal or lower variance. Two factors account for the discrepancy. First, positive-pair density per batch controls how often saturation occurs. Second, task difficulty compresses seed-to-seed spread when accuracy is high. An SVHN difficulty sweep confirms the interaction on a single dataset, with the variance ratio moving from $0.25\times$ at high accuracy to $16.73\times$ under aggressive augmentation. In moderate-accuracy regimes with many same-class pairs per batch, switching to the gradient-neutral subtraction reference removes this variance inflation at no cost to mean accuracy. Measuring the layer-0 clamp activation rate serves as a simple check for whether the problem applies.

cs.LG

PAC Guarantees for Reinforcement Learning: Sample Complexity, Coverage, and Structure

When data is scarce or mistakes are costly, average-case metrics fall short. What a practitioner needs is a guarantee: with probability at least $1-\delta$, the learned policy is $\varepsilon$-close to optimal after $N$ episodes. This is the PAC promise, and between 2018 and 2025 the RL theory community made striking progress on when such promises can be kept. We survey that progress. Our organizing tool is the Coverage-Structure-Objective (CSO) framework, proposed here, which decomposes nearly every PAC sample complexity result into three factors: coverage (how data were obtained), structure (intrinsic MDP or function-class complexity), and objective (what the learner must deliver). CSO is not a theorem but an interpretive template that identifies bottlenecks and makes cross-setting comparison immediate. The technical core covers tight tabular baselines and the uniform-PAC bridge to regret; structural complexity measures (Bellman rank, witness rank, Bellman-Eluder dimension) governing learnability with function approximation; results for linear, kernel/NTK, and low-rank models; reward-free exploration as upfront coverage investment; and pessimistic offline RL where inherited coverage is the binding constraint. We provide practitioner tools: rate lookup tables indexed by CSO coordinates, Bellman residual diagnostics, coverage estimation with deployment gates, and per-episode policy certificates. A final section catalogs open problems, separating near-term targets from frontier questions where coverage, structure, and computation tangle in ways current theory cannot resolve.

cs.LG

PyHealth 2.0: A Comprehensive Open-Source Toolkit for Accessible and Reproducible Clinical Deep Learning

Difficulty replicating baselines, high computational costs, and required domain expertise create persistent barriers to clinical AI research. To address these challenges, we introduce PyHealth 2.0, an enhanced clinical deep learning toolkit that enables predictive modeling in as few as 7 lines of code. PyHealth 2.0 offers three key contributions: (1) a comprehensive toolkit addressing reproducibility and compatibility challenges by unifying 15+ datasets, 20+ clinical tasks, 25+ models, 5+ interpretability methods, and uncertainty quantification including conformal prediction within a single framework that supports diverse clinical data modalities - signals, imaging, and electronic health records - with translation of 5+ medical coding standards; (2) accessibility-focused design accommodating multimodal data and diverse computational resources with up to 39x faster processing and 20x lower memory usage, enabling work from 16GB laptops to production systems; and (3) an active open-source community of 400+ members lowering domain expertise barriers through extensive documentation, reproducible research contributions, and collaborations with academic health systems and industry partners, including multi-language support via RHealth. PyHealth 2.0 establishes an open-source foundation and community advancing accessible, reproducible healthcare AI. Available at pip install pyhealth.

cs.LG

Spectral and Homological Bounds on k-Component Edge Connectivity

We present a novel theoretical framework connecting k-component edge connectivity with spectral graph theory and homology theory to pro vide new insights into the resilience of real-world networks. By extending classical edge connectivity to higher-dimensional simplicial complexes, we derive tight spectral-homological bounds on the minimum number of edges that must be removed to ensure that all remaining components in the graph have size less than k. These bounds relate the spectra of graph and simplicial Laplacians to topological invariants from homology, establishing a multi-dimensional measure of network robustness. Our framework improves the understanding of network resilience in critical systems such as the Western U.S. power grid and European rail network, and we extend our analysis to random graphs and expander graphs to demonstrate the broad applicability of the method. Keywords: k-component edge connectivity, spectral graph theory, homology, simplicial complexes, network resilience, Betti numbers, algebraic connectivity, random graphs, expander graphs, infrastructure systems

math.CO

On the Degree Sequences of Multigraphs with Edge Additions and Deletions

The degree sequence of a graph is a numerical method to characterize the properties of graphs. Generalized forms of degree sequences exist for complete graphs and complete graphs. Nikolopolus et al. characterized the number of spanning trees from edge deletions and edge additions. Instead of investigating the number of spanning trees of graphs that arise from edge additions and deletions, we sought to characterize degree sequences of such graphs. We conjecture a characterization for the degree sequence of the addition and edge deletion for many families of graphs including threshold graphs and complete multigraphs. Keywords: multigraphs, split graphs, degree sequence, threshold graph, Havel-Hakimi, Ruch-Gutman, Edge Deletion

math.CO

A Characterization of Distance Matrices of Positive Weighted Kneser Graphs and Generalized Petersen Graphs

Rubei et. al., established results for the distance matrix of positive weighted Petersen graphs. Focusing on the properties of the distance matrix, we generalized positive weighted Petersen graphs results to Kneser graphs. We analyzed theorems established by Rubei et al. and used girth of the generalized Petersen graphs and Kneser graphs to conclude generalizations. Further, we examined the properties of positive weighted generalized Petersen graphs. We generalized the properties of distance matrices of positive weighted Petersen graphs to positive weighted generalized Petersen graphs.

math.CO