Commutative Algebra
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Friday 6 November
If $S$ is a numerical semigroup, then we denote by $m(S)$, $F(S)$ and $\mathrm{PF}(S)$ its multiplicity, its Frobenius number and its set of pseudo-Frobenius numbers, respectively. Let $\mathcal P$ be a finite nonempty set of positive integers. In this paper, we prove that $\mathcal C(\mathcal P)=\{S\mid F(S)=\max(\mathcal P),\ \mathcal P\subseteq\mathrm{PF}(S)\}$ is a covariety. The associated tree provides algorithms to compute $\mathcal C(\mathcal P)$, its elements with a prescribed type or genus, and the set of numerical semigroups $S$ satisfying $\mathrm{PF}(S)=\mathcal P$. We show that every element of $\mathcal C(\mathcal P)$ has a unique minimal system of generators relative to $\mathcal C(\mathcal P)$, and we call its cardinality the $\mathcal C(\mathcal P)$-rank. Finally, we characterize the elements of $\mathcal C(\mathcal P)$ with $\mathcal C(\mathcal P)$-rank one.
Machine Learning
MINE: Entropy-coded Memory-efficient Inference for Large Language Models
We present MINE (Memory-efficient INference Engine), a lossless entropy coding approach that accelerates single-stream inference of Large Language Models (LLMs), where existing lossless compression methods yield speedups only for specific data types such as BF16 and under additional assumptions. Single-stream decoding is strongly memory-bound, since every weight must be streamed from memory once per generated token and requests cannot be batched in local or low-latency deployments. At the core of MINE is dtANS-LLM, a GPU entropy coding scheme adapted from the prior dtANS to the needs of LLM inference. It exploits the non-uniform distribution of weights, which persists even after quantization, and exposes hyper-parameters that trade compression ratio against decoding cost. We auto-tune these parameters for each tensor and GPU to obtain the Pareto front between model size and decoding latency. MINE integrates into vLLM as a plugin and selects the best kernel configuration for a given memory budget. On Llama 3.1 8B Instruct with a 2.1-bit EntQuant checkpoint, MINE achieves speedups of 2.52x and 1.48x over the fastest baseline on an RTX 4060 Ti and an RTX 5090, at effective bitrates of 2.24 and 3.43 bits per weight. MINE is available as open source.
Read the full abstract
Artificial Intelligence
Agentic AI for Healthcare Claim Denial Analysis: A Policy-Grounded Framework for Privacy Preservation
In the healthcare sector, claim denials impose significant financial and administrative burdens, emphasizing the need for automated and trustworthy decision-support systems. This paper proposes a modular Agentic Artificial Intelligence (AI) framework for healthcare claim denial analysis, integrating specialized agents for denial risk prediction, risk factor identification, diagnosis-procedure-medication code relationship analysis, and counterfactual analysis. The framework integrates predictive analytics, graph-based learning, a Hybrid Contextual Retrieval-Augmented Generation (RAG) pipeline grounded in healthcare policy documents, and a dedicated HIPAA Compliance Agent to support privacy protection. Evaluated on a synthetic healthcare claims dataset, the framework achieved 98% traceability of reported numerical findings, validated 214 of 216 policy citations, and removed all 280 identified sensitive identifiers without modifying the underlying analysis. These findings pave the way for further research into Agentic AI as a framework for explainable, policy-aware, and privacy-preserving healthcare claim denial management.
Read the full abstract
Cryptography and Security
Robust & Scalable Website Fingerprinting via Minimal DRAM Traffic
Website fingerprinting attacks infer websites visited by users and threaten their privacy. These attacks are classified as on-path, which monitor network traffic, and co-located, which exploit shared system resources. Existing co-located attacks, such as sweep-counting and loop-counting, achieve over 90% accuracy but suffer significant performance degradation on many-core systems. Loop-counting accuracy drops from 93% on 8 cores to 50% on 24 cores, while sweep-counting accuracy decreases from 84% to 45% across multiple Core Complex Dies (CCDs). To address these limitations, we propose memory-chasing, a scalable website fingerprinting attack that exploits DRAM traffic interference through pointer-chasing traversals with minimal interference. By counting DRAM accesses over fixed intervals, the attack generates distinctive traces for website classification. Evaluated on the top 50 websites, memory-chasing achieves 96% accuracy on a 24-core system, 86% across multiple CCDs, 80% across multiple sockets, and 97% in an open-world setting. Extensive evaluation across hardware configurations, browsers, operating systems, isolation mechanisms, system noise, and countermeasures demonstrates its robustness and scalability.
Read the full abstract
Data Structures and Algorithms
Improved Approximation and Hardness Bounds for $\beta$-TSP
The traveling salesman problem (TSP) asks for a minimum-length Hamiltonian cycle in a complete weighted graph. Its approximability depends on the edge lengths: unrestricted instances admit no constant-factor approximation unless $\mathrm{P}=\mathrm{NP}$, metric instances admit Christofides' $3/2$-approximation, and fixed-dimensional Euclidean instances admit a PTAS. We study $\beta$-TSP, where the edge lengths satisfy $ \ell_{uw}\leq\beta(\ell_{uv}+\ell_{vw}) \qquad \forall\,u,v,w\in V\ \text{pairwise distinct}, \quad \beta\in\left[\tfrac{1}{2},1\right]. $ The condition arises naturally in various settings, for example, when metric travel times are combined with positive service times at the vertices. The previous best approximation algorithm, due to Böckenhauer et al., ranges from $1$ at $\beta=1/2$ to $3/2$ as $\beta$ approaches $1$. We obtain a strictly better guarantee for every $1/2<\beta<1$ by combining two simple constructions whose guarantees counterbalance each other. We also improve the best known approximation hardness bound for every $1/2<\beta<1$.
Read the full abstract
High Energy Physics (Theory)
Null Strings Gauged, Reloaded and Quantized, II: Null Strings on Flat Backgorund Have Discrete Spectrum
We continue our study of the quantization of null strings, strings with Carrollian worldsheets described by an action with two-dimensional (2d) diffeomorphisms, Weyl and Carroll-Weyl scaling as gauge symmetry. We perform light cone gauge quantization of the null strings on flat space background within Schrödinger representation, initiated in arXiv:2607.02970. We show, within the regular separated Schrödinger sector, that bosonic null strings on a flat background have no critical dimensions, have discrete, nonnegative mass-squared levels and are hence free of tachyonic modes.
Read the full abstract
Mathematical Physics
Zak Phase, Pole–Zero Invariants and the Maslov Index of Bloch Cauchy Lines in One-Dimensional Periodic Media
The topological properties of the Bloch bands of a one-dimensional periodic medium are described either by the Berry–Zak phase or by the poles and zeros of the surface impedance. We show that both can be read from one geometric object: the line spanned by the field and its flux (the Cauchy data) at an inversion centre. For each band, this line turns as the quasi-momentum runs over the Brillouin zone. Its number of turns, the Maslov index, is fixed by the parities (even or odd) of the Bloch functions at the two band edges, and its parity is the Zak phase. In a gap, the Cauchy line of the solution decaying on one side rotates monotonically with the energy, and an interface state between two crystals is a crossing of two such lines. This gives the known criterion of at most one state per gap in terms of the same lines. Examples with a stratified medium show that the choice of inversion centre changes the interface spectrum. The last section discusses what is missing to extend the construction to higher dimension.
Read the full abstract
Human-Computer Interaction
DeepRadio: A Human-Agent Workspace for Physical Radio Development
Agentic programming can automate software-defined radio (SDR) development, but decisions can propagate from software to hardware without sufficient user visibility or control. We present DeepRadio, a human-agent workspace embedded in GNU Radio Companion (GRC) for jointly constructing, verifying, and operating physical radio systems. DeepRadio uses a dynamic and state-driven workflow to coordinate a traceable radio specification, claim-oriented cross-layer verification, and mixed-initiative control. These mechanisms expose decisions, tie conclusions to available evidence, and return control to users before consequential hardware actions. Across seven atomic SDR tasks, DeepRadio completed nearly all runs while satisfying all prescribed software and radio-frequency constraints. In a 20-participant study, more participants achieved a verified physical deployment with DeepRadio than with Codex-assisted GRC. Participants reported stronger support for shared intent and understanding while retaining initiative and control. These results suggest that DeepRadio can make agent-assisted physical radio development more reliable, understandable, and controllable.
Read the full abstract
Machine Learning
Future-Informed Actor-Critic
Actor-critic methods optimize a parametrized policy using a critic to estimate their policy gradient. Recent work has shown that conditioning the critic on features of future states, actions, or rewards can improve sample efficiency and even final return. However, existing methods are restricted to exogenous features (independent of actions), overlooking endogenous features (dependent on actions) that may be crucial for credit assignment. We introduce the future-informed value function, which can be conditioned on both exogenous and endogenous features of future states, actions, and rewards. Then, we express the policy gradient using this value function, yielding the future-informed policy gradient. Building on that, we propose the future-informed actor-critic algorithm. Empirical results show that this algorithm leveraging endogenous features of the future outperforms standard actor-critic algorithms in environments with noisy and sparse rewards.
Read the full abstract
Combinatorics
A 4-regular graph with 4/5 of its vertices satisfying the Šoltés property
The Wiener index $W(G)$ of a connected graph $G$ is the sum of distances between all unordered pairs of its vertices. A vertex $v$ is called a Šoltés vertex if $G-v$ is connected and $W(G-v)=W(G)$. Since the cycle $C_{11}$ is the only known graph in which every vertex is a Šoltés vertex, a natural problem is to construct other graphs with as large a proportion of such vertices as possible. We construct a $4$-regular graph with Šoltés fraction $4/5$, improving the previous value of $2/3$.
Read the full abstract
Rings and Algebras
Inclusions and intersections of products of verbally prime $T$-ideals
We study inclusions and intersections of products of verbally prime $T$-ideals over a field of characteristic zero. Our main tool is a Vandermonde-type construction of graded polynomials with a uniformly bounded number of variables. Together with a reduction to finite exterior algebras, this construction yields separating polynomials that work uniformly under bounds on Grassmann envelope representations. We prove that, under these bounds, an intersection of $T$-ideals is contained in such a product only if one member is. This proves Berele's Conjecture~3 and, consequently, his Conjecture~1 \cite{Berele2012}. We also characterize inclusions between ordered products by an explicit combinatorial criterion, proving his Conjecture~2 and answering his Question~2 \cite{Berele2012}. These results characterize inclusions between arbitrary intersections of products of nonzero proper verbally prime $T$-ideals and show that every nonzero such intersection has a unique finite irredundant representation. As an application, we characterize finite quivers whose path algebras have the same polynomial identities as some, equivalently every, positive power of their arrow ideals.
Read the full abstract
Artificial Intelligence
Typed-Centrality Grounding for Agentic GraphRAG: A Symbolic Approach for Agentic Reasoning
Knowledge Graphs (KGs) and Large Language Models (LLMs) represent two approaches for knowledge representation and reasoning over heterogeneous data. Recent research has explored their integration through Graph Retrieval-Augmented Generation (GraphRAG) systems, which use KGs to provide structured context to LLMs. However, existing GraphRAG approaches treat KGs as homogeneous graphs and rely on textual similarity between the input prompt and the context, rather than using it as a source of deterministic knowledge. This work proposes an agent-based GraphRAG system that introduces a symbolic layer to integrate deterministic reasoning alongside text similarities. This approach is based on the definition of typed centralities for KGs, where centrality measures are computed separately for each specific KG's class. This information is used to construct a symbolic representation that is provided to the LLM as a deterministic source. The proposed system was mathematically validated and evaluated on a real-world POLE (Person, Object, Location, Event) dataset, comparing reasoning and information retrieval with and without the symbolic layer. The results indicate that the our approach leads to responses that are more consistent with the context, improving the generated outputs.
Read the full abstract
Medical Physics
SmartBloch-Agent: An Agentic Framework for Automated MRI Sequence Design and Optimization
Magnetic resonance imaging (MRI) pulse sequences organize radiofrequency excitation, gradient encoding, and signal acquisition, determining magnetization evolution and image formation. Sequence research requires translating imaging objectives into executable protocols meeting physical and acquisition constraints, refined via physical simulation, signal reconstruction, and scientific evaluation. Sequence construction, physical verification, parameter search, and assessment involve many tools and manual decisions, limiting automation from objective to verifiable sequence. We develop SmartBloch-Agent, an integrated framework for automated MRI sequence design and optimization. Given objectives and scientific constraints in natural language, an AI agent interprets goals, plans tasks, orchestrates tools, and makes feedback-informed decisions, drawing on a construction module and optimizers for structural design and parameter search. Our GPU-accelerated SmartBloch engine supplies physical computation for both stages, so the framework configures, per sequence and objective, event validation, Full-Bloch magnetization evolution, signal acquisition, image reconstruction, and scientific evaluation, and closes the loop without external MRI simulators. Results return to the agent to guide sequence revision, parameter optimization, and candidate selection, and the design files carry simulation, reconstruction, and evaluation evidence, forming an output package for a real-scanner workflow. Experiments show the framework automatically designs and optimizes MRI sequences for varied objectives and constraints, refines candidates with physical feedback, and improves several parameter-estimation metrics and downstream quantitative reconstruction. SmartBloch-Agent unifies objectives, design, physical execution, and evaluation into a continuous workflow for end-to-end automation of MRI sequence research.
Read the full abstract
Artificial Intelligence
MIRA: Mitigating Pole Geometry Bias in Recurrent Language Models
We show that the Transformer's recurrent successors — such as selective state-space models (SSMs), Receptance Weighted Key Value (RWKV), and Retentive Network (RetNet) — can be characterized as infinite-impulse-response (IIR) systems. The zero–pole geometries of these IIR systems characterize the temporal modes they can represent. However, to trade parallelism, the algebraic structures of the parameterizations of these recurrent systems often give rise to poles constrained to the real axis, thereby inducing an architectural bias in their recurrent dynamics. To unify these recurrent architectures and eliminate their pole geometry bias, inspired by the lattice–ladder realization of general IIR systems, we present Mira (Modal IIR Recurrent Architecture), a unified family of linear-time recurrent architectures for building large language models (LLMs) with a highly parallel algebraic structure. Each Mira block realizes a higher-order IIR system that represents temporal modes from token streams through a lattice–ladder structure, followed by a feed-forward network (FFN). This realization enables the model to represent the temporal modes associated with both real and complex-conjugate poles. We pretrain Mira-S, a 12-layer small-scale LLM with 0.14B parameters composed entirely of Mira blocks. Throughout the pretraining, Mira-S achieves lower perplexity than strong recurrent baselines, exhibits stronger context-window generalization beyond the training length, and attains higher throughput with a dedicated CUDA kernel.
Read the full abstract
Distributed Computing
DataHelix: A CPU-GPU Co-Scheduling Data Plane for Multimodal Model Training
Multimodal model training breaks the loose coupling between data loading and model execution assumed by conventional training pipelines. Large per-sample variation in decoding cost, token/latent count, and tensor size causes both CPU preprocessing time and GPU compute time to vary substantially across data-parallel ranks (training processes). In synchronous training, these two sources of heterogeneity interact tightly: GPU execution determines how long the input pipeline has to prepare the next batch, while delayed preprocessing in any rank in turn stalls all GPUs at synchronization. As a result, optimizing CPU throughput or GPU load balance in isolation cannot eliminate pipeline bubbles. We present DataHelix, a metadata-first CPU-GPU co-scheduling data plane for multimodal training. DataHelix jointly reasons about preprocessing demand and GPU execution progress, treating each GPU compute window as the readiness budget for the next batch and coordinating preprocessing across ranks to minimize deadline misses rather than CPU makespan. To avoid expensive data movement, placement is decided while samples remain lightweight descriptors, before decoding and collation amplify them into large tensors. Descriptor-level token-latent-aware batching performs coarse pre-balancing, while cost-gated collective rebalancing corrects residual skew only when its expected benefit exceeds transfer cost. Across 11 video and multimodal workloads on 8 Ascend 910B accelerators, DataHelix reduces mean deadline-miss rate by 67.8% and mean p95 input-wait latency by 87.0%, while achieving 2.72x geometric-mean throughput over rank-local DataLoaders and 1.82x over a Ray actor pool under the same CPU budget.
Read the full abstract
Programming Languages
Nothing Is Shared Until You Synchronise: Reinterpreting Java for Data-Race Freedom
In mainstream shared-memory languages, programmers use synchronisation mechanisms such as locks to isolate shared data accesses across threads to avoid data races. Yet synchronisation enforces only a temporal ordering, while shared memory still communicates updates between threads at every ordinary read and write. To achieve isolation, programmers therefore need to understand the connection between ordering and communication, and to apply synchronisation correctly and consistently, while compilers and hardware optimise assuming they have. When programmers synchronise incorrectly, they lose isolation, and the semantics of their programs change without notice. We propose Sync-as-Share, a memory model built on isolation and communication rather than on the ordering of accesses to shared memory. We start from thread isolation, where each thread reads and writes a private view of memory, and reinterpret programs' existing synchronisation actions to also communicate updates explicitly. As nothing is communicated outside synchronisation actions, data races become impossible by construction. Instead, conflicting updates are detected at synchronisation points. Viewing synchronisation as explicit communication exposes the design decisions of what should be communicated, to whom, and what counts as a conflict. We propose two communication topologies with different trade-offs. We instantiate Sync-as-Share for Java by making the synchronisation actions already defined by its memory model the only points at which threads communicate, so Java programs can adopt Sync-as-Share unmodified. We formalise this instantiation as version-controlled memory, in which synchronisation actions push and pull updates between private views, and mechanise it in Rocq. We show that Java programs that are correctly synchronised under the Java memory model behave identically under shared memory and Sync-as-Share.
Read the full abstract
Signal Processing
Beyond the Pseudo-Inverse: GIM Precoding for Per-Antenna Power Minimisation in MU-MIMO
We study linear precoder design for multi-user MIMO downlink under per-antenna power constraints (PAPC), which arise naturally when each antenna port has its own power amplifier. While the Moore-Penrose pseudo-inverse is the unique optimal generalized inverse for zero-forcing precoding under a total power constraint [1], we prove this optimality does not extend to the MMSE/RZF case under PAPC. Any precoder in the Generalized Inverse Matrix (GIM) family shares the same effective channel and per-user SINRs as the pseudo-inverse baseline, while its null-space component reshapes the per-antenna power profile. This separation principle enables minimising peak per-antenna power at fixed SINR targets with zero SINR penalty. We characterise the convex peak-power problem over this family and its active-set optimality, establish MRT and ZF asymptotic limits, and provide a convergent first-order subgradient algorithm. Simulations with 8-64 antennas confirm gains of 1.9-3.7 dB over pseudo-inverse RZF, with GIM-MMSE eliminating essentially all PAPC violations up to 20% per-antenna budget imbalance. The framework applies directly to fully-digital massive MIMO. Cell-free and hybrid analog-digital arrays are promising extensions.
Read the full abstract
High Energy Physics (Phenomenology)
Parton density functions for future electron-positron colliders
Precision QED radiative corrections are critical for the physics programs of future high-energy $e^+e^-$ colliders. We review the state-of-the-art QED parton density function (PDF) formalism for resumming collinear photon and pair emissions. Our numerical analysis of initial-state radiation (ISR) highlights the slow convergence of the perturbative series and the growing significance of pair production at higher orders. Finally, we emphasize the need to implement these theoretical refinements into modern Monte Carlo generators to meet the stringent precision demands of upcoming experimental measurements.
Read the full abstract
Artificial Intelligence
ROAM: Retrieval-Augmented Optimization for Feasible Multi-Day Itinerary Planning
Large language models (LLMs) can draft plausible travel itineraries, but the plans they produce are frequently infeasible: attractions are grouped into days without regard for geography, opening hours are ignored, and a single generation pass cannot reliably organize a multi-day trip. We present ROAM, a retrieval-augmented agentic framework for spontaneous multi-day itinerary planning that couples semantic preference retrieval with explicit spatiotemporal optimization. ROAM embeds the natural-language request, retrieves a compact candidate set from a curated point-of-interest (POI) database, partitions candidates into day groups through maximal-clique analysis and K-means, orders each day with a quality-biased simulated-annealing TSP solver, and repairs hard opening-hour violations with greedy exchange. A spatiotemporal retrieval layer injects weather forecasts, real road distances and POI operating hours into the prompt, after which the LLM renders a structured, map-ready itinerary; an interactive ReAct-style replanning agent then supports natural-language edits while re-checking feasibility. Across four Chinese cities and 510 POIs, ROAM produces 27.5 POIs per three-day plan versus 19.3 for a bare LLM and 17.0 for a retrieval-only baseline (+42% and +62%), while reducing route dispersion from 2,101 to 677 (3.1x). Ablations show that spatial optimization is the dominant driver of route efficiency and that hard time-window repair prevents roughly 30% of attractions from violating operating hours.
Read the full abstract
Quantum Physics
Quantum-induced latent symmetries
We study the interplay of latent symmetries and multiparticle quantum states upon the introduction of additional particles. We show how physical symmetries in multimode quantum systems give rise to symmetries in the multiparticle state space of indistinguishable fermions and bosons. Latent symmetries, hidden symmetries that are exclusively visible in the eigenspectral domain, offer new ways to engineer the functionality of multimode quantum systems. In this vein, we present multiboson state generation as an application of latent symmetries to the multiparticle domain. Finally, we demonstrate that even in systems that are entirely devoid of symmetry, latent symmetries can be induced in the multiparticle state space, as a result of quantum indistinguishability.
Read the full abstract
Quantum Physics
Flexible Control of Temporal Reflection and Refraction via Joint Parameter Modulation
Achieving the flexible control of wave propagation is a central goal in the design of optical devices. In this work, our analytical results and numerical simulations show that in both the two-legged model and SSH model, joint modulation of coupling strengths and phases enables independent control of the probability partition and the propagation opening angles in the space-time map. We hope these results could provide a theoretical foundation for the design of programmable time-modulated optical devices.
Read the full abstract
Nuclear Theory
Effect of collective nuclear vibrations on initial anisotropies in central Pb+Pb collisions: A comparative study in Glauber and Trento models
We perform a comparison of the Glauber and Trento model predictions for the initial state azimuthal anisotropies $\epsilon_{2,3}\{2\}$ in central Pb+Pb collisions at the LHC energies. We incorporate the quantum collective effects in $^{208}$Pb by sampling nucleon positions with filterings in the quadrupole and octupole moments. We use the data-driven filters adjusted to reproduce the mean squared quadrupole and octupole moments extracted from the experimental strength functions. We find that the collective effects reduce considerably the $\epsilon_2\{2\}/\epsilon_3\{2\}$ ratio. The reduction is more pronounced in the Trento model. We obtain $\epsilon_2\{2\}/\epsilon_3\{2\}\approx 0.75\div 0.79(0.68\div 0.71)$ in the Glauber(Trento) models. These values allow to resolve the $v_2$-to-$v_3$ puzzle.
Read the full abstract
Computers and Society
Uniqueness-Aware Feature Engineering for Representing Humans in Algorithmic Decision Making
Fairness analyses of algorithmic decision systems typically treat the feature representation of individuals as an a priori fixed input. How well a feature representation captures what is actually decision-relevant for different people and their unique backgrounds, however, is typically left unexamined. We address this gap by modeling feature engineering as a stochastic selection process over the pooled characteristics as candidate features for a finite training population. Based on this model, we develop a new framework for uniqueness-aware feature engineering which operates in two layers. The first layer measures population heterogeneity using a multivariate hypergeometric model of feature prevalence. It shows how well an individual is represented by any given input representation and which individual of the population is "least represented" compared to their personal, reference characteristics. The second layer acts on this assessment using set-similarity measures to search for an optimum between representational fit of the population and predictive accuracy. Without assuming prior knowledge about unseen individuals, our framework grants represented humans agency over what counts as decision-relevant in their case, gives deploying institutions accountability through a publishable representation fit report, and provides each person subject to a decision transparency about their personal representation fit. We operationalize our framework as a six-step pipeline and show a proof-of-principle using synthetic data. Our approach offers a blueprint for making algorithmic decision systems accountable for how well they represent individuals during development and in production.
Read the full abstract
Computation and Language
A Spiking Neural Model of the Mental Lexicon: Representations, Similarity Structure, and Associative Retrieval in Perception and Production
A neural account of lexical access requires an explicit distinction between the structure of lexical representations, the associations linking representational levels, and the activity generated during retrieval. This paper specifies a spiking neural implementation of the mental lexicon within the Neural Engineering Framework and Semantic Pointer Architecture. Conceptual and phonological form representations are constructed recursively from relational graphs, whereas lemma representations provide independent lexical identities. The implementation contains 613 lexical entries and supports two directed processing pathways: perception maps a phonological form to its associated concept via a lemma, whereas production maps a concept to its associated phonological form via a lemma. Each pathway combines two associative-memory variants, winner-take-all (WTA) selection (competition) and thresholded transfer, and uses 199,400 leaky integrate-and-fire neurons. We formalize within-level similarity, cross-level associative retrieval, and the interpretation of decoded activity. Simulations show successful retrieval of eight selected words from the lexicon in both processing directions, with retrieval evaluated against all 613 lexical entries. These simulations use complete phonological forms as input for perception and complete concepts as input for production, illustrating lexical retrieval through the corresponding processing pathways. This paper provides a transparent description of lexical representations and neural retrieval mechanisms, supported by accessible code, lexical data, and online simulation documentation.
Read the full abstract
Artificial Intelligence
PhyCAD: Physics-Aware Text-to-CAD Generation via Hierarchical Reinforcement Learning
Language-driven computer-aided design (CAD) generation enables the creation of editable parametric models from text, yet geometric correctness alone does not ensure physical feasibility. We present PhyCAD, a framework for physics-aware Text-to-CAD generation that trains language models to produce executable CadQuery programs under geometric and physical constraints. To support this setting, we construct a Text-to-CAD dataset that links parametric CAD models with geometric specifications, loading conditions, and physical constraints. We further develop an automated CAD–CAE verifier that combines program execution, geometric checks, and finite element analysis to provide feedback for model training. A key learning challenge is to prioritize joint satisfaction of geometric and physical constraints while retaining informative feedback on partial progress. A flat aggregation of heterogeneous rewards can obscure this priority, allowing improvements in individual criteria to compensate for unmet design requirements. We therefore introduce a Hierarchical Balanced Advantage for Group Relative Policy Optimization, grouping candidates into levels according to their verification status and combining between-level progress with objective-wise peer ranking within each level. This formulation preserves cross-level priority in the training signal while providing fine-grained feedback for satisfying constraints and improving feasible designs. Experiments across multiple language models show that PhyCAD consistently improves joint geometric and physical constraint satisfaction. These results demonstrate the value of verifiable engineering feedback for advancing language-driven CAD generation beyond geometric reconstruction toward physics-aware design.
Read the full abstract
Machine Learning
Causal graph fusion across incomplete cohorts and causal steering of tabular foundation models
Clinical data are split across incomplete cohorts with partially overlapping variables. Causal discovery under missing data and multi-source integration have developed separately, and tabular foundation models cannot take a causal graph without retraining. We make two contributions. First, causal graph fusion learns a partial ancestral graph on the shared variables, fuses it into each cohort's graph under temporal background knowledge and lets Markov blankets guide re-imputation, so missing data are handled inside discovery. Second, causal attention steering compiles the learned graph into the attention of a frozen tabular foundation model at inference. Against known ground truth, fusion dominates pooled and per-cohort baselines in every missingness and heterogeneity condition tested (p<0.001), and steering improves discrimination under distribution shift (p<10^{-5}), a gain that a random graph does not produce. In two spinal cord injury cohorts, fusion orients age, albumin and hemoglobin as ancestors of length of stay, a set no single cohort orients in full, and yields compact, temporally ordered predictor sets that match or exceed automatic feature selection on four of five outcomes.
Read the full abstract
Theoretical Economics
Artificial Shoppers and Attention Competition
We study how AI shopping interfaces affect price competition, platform profit, and consumer welfare. A platform hosts two of three products of different quality. Without AI, consumers compare the hosted pair under salience distortion and overweight whichever attribute, price or quality, stands out. Under assistance, an AI assistant recommends the two products with the highest true utility, leaving consumers to make the same distorted comparison. Under delegation, an AI agent selects the best product, subject to purchase approval. Both modes access all products and know consumers' preferences; firms price strategically. First, assistance can soften competition despite expanded product access. The middle-quality firm may forgo low-valuation consumers because attracting them requires cutting the price paid by higher-valuation consumers as well. Relative to no AI with the same assortment, assistance lowers both hosted prices under the high- and low-quality pair but can raise both under the high- and middle-quality pair. Second, we identify an alignment paradox: some consumers can become worse off than without AI even though their best option remains recommended and becomes cheaper. Salience can induce a switch to a lower-quality product whose quality loss exceeds the saving relative to the original purchase. The assistant ranks products accurately but does not optimize the shortlist for consumers' realized welfare. Third, assistance can raise both platform profit and aggregate consumer surplus with the high- and middle-quality pair; delegation can do so with the high- and low-quality pair. Fourth, among AI designs, the platform weakly prefers delegation with the high- and middle-quality pair wherever all compared pricing equilibria exist. Yet this preferred design can yield lower platform profit and aggregate consumer surplus than the no-AI optimum.
Read the full abstract
High Energy Physics (Theory)
Towards a recursive construction of string scattering amplitudes
We construct open superstring scattering amplitudes from three-point data using the Britto-Cachazo-Feng-Witten (BCFW) recursion and four-dimensional spinor helicity variables. We compute the contributions of selected leading and subleading Regge families to the color-ordered four-gluon amplitude by sewing their cubic couplings and summing over intermediate polarizations. The four-dimensional description makes helicity selection rules explicit and reduces the massive polarization sums to SU(2) little group contractions. The resulting residues are expressed in terms of Jacobi and Legendre polynomials, yielding explicit sums over mass levels. We construct a differential operator relating the two polynomial kernels, including their level-dependent normalizations. These results provide a concrete starting point for organizing the infinite intermediate-state sum and investigating whether relations among Regge families can support a recursive reconstruction of the full string amplitude.
Read the full abstract
Astrophysics of Galaxies
Where the red rise stops: JWST/MIRI reveals a near-infrared turnover in the Little Red Dot Abell2744-QSO1
Little Red Dots (LRDs) are compact sources with a blue rest-frame UV, a steep red optical continuum and, often, broad hydrogen lines. They have been interpreted as dense stellar systems, dust-reddened active galactic nuclei, or black holes embedded in dense, dust-poor gas, scenarios that are hard to tell apart in the rest-frame UV and optical. We present JWST/MIRI F770W and F1000W imaging of the triply lensed Abell2744-QSO1 at z=7.04, probing rest-frame 0.96 and 1.24um. QSO1 is unresolved, and its steep optical rise stops near rest-frame 1um. Up to 1.24um, a T~4450K blackbody, or a hotter one reddened by dust, describes the continuum, with no sign of the hot dust seen in most type-1 quasars. Our statistically preferred SED fits produce the red continuum with (3-4)x10^9 Msun of stars, over 100 times the dynamical limit on the host mass, and are therefore disfavoured. An accretion disc in a dense gas cocoon respects this limit but over-predicts the 1.24um flux by 5 sigma. Super-Eddington accretion models with a thermalising envelope reproduce the turnover, but only for a black hole ~2.5dex less massive than the dynamical estimate. For a dust-reddened accretion disc (A_V ~ 2 mag), dust at ~1200K can re-emit at most ~2 per cent of the absorbed luminosity, and most of it would have to emerge at ~70-600K, unless the dust covers only a small fraction of the sky. The dynamics favour an accretion-powered source, but our data cannot distinguish dust reddening from processing by dense gas. MIRI imaging at >18um could separate these scenarios.
Read the full abstract
Artificial Intelligence
Travel Cost Calibration for Multi-Agent Target Assignment and Pathfinding
Target assignment and pathfinding (TAPF) is an appealing extension of classical multi-agent pathfinding (MAPF): each agent has a set of targets, and the planner must both assign a target to each agent and plan collision-free paths for them. A trivial approach is to compute an assignment that minimizes the total single-agent shortest-path distance over agent–target pairs, and then solve the resulting MAPF instance. However, this approach falls short because the shortest-path distance remains a proxy that ignores interference among agents. Assigning targets based on the costs agents actually pay, which depend on how every other agent is assigned and routed, would yield a better final outcome. To this end, we propose an iterative TAPF method that improves target assignments by calibrating travel cost estimates from fast, suboptimal MAPF feedback. Experiments show that our method produces solutions close to optimal on small instances and outperforms existing planners in solution quality on larger ones.