Representation Theory
Central algebraic cohomology of Galois gerbes
Taeyeoup Kang
We study central algebraic
cohomology of Galois gerbes over local fields of characteristic zero and number fields, with coefficients in a connected reductive group and an algebraic central subgroup, not necessarily finite or connected. For gerbes admitting a comparison morphism from Kaletha's rigid gerbe, we obtain fiber product formulas for abelianized
cohomology in terms of algebraic fundamental groups of finite central quotients and homomorphisms from the band to the central subgroup. Such comparisons exist whenever the band is of finite type. Over number fields, independently of the comparison hypothesis, the full
cohomology set is recovered from its abelianization and its real localizations through a cartesian square. Over non-Archimedean local fields and totally imaginary number fields, abelianization is bijective. We also specialize the formulas to Kottwitz's local and global gerbes.
Read the full abstract ▾
Computers and Society
Understanding Student Use of Large Language Models Across Computer Science Subfields
Sehrish Basir Nizamani, Yoonje Lee, Nikitha Donekal Chandrashekar et al. (+2 more)
This research full paper examines how undergraduate students use large
language models (LLMs) across computer science subfields. As LLMs become increasingly integrated into computing education, understanding how their use varies across technical and pedagogical contexts is essential for designing effective, subfield-aware instruction. This paper presents a cross-subfield analysis of LLM usage among 211 undergraduate students in a problem-solving course intentionally designed to support responsible and effective LLM use through structured instruction and reflection. Using post-assignment reflection data collected across seven instructional modules spanning multiple computer science subfields, we examine
prompt counts, LLM role conceptualization, and verification behavior. Results show that LLM adoption varies substantially by assignment, with higher usage in
algorithms and web development and lower usage in software engineering. Students predominantly treat LLMs as assistive tools rather than authoritative sources, and verification is common across all subfields, with most students using multiple strategies. Verification behavior also varies by assignment context, with testing more common in structured tasks and web search more common in open-ended tasks. These findings suggest that assignment characteristics play a central role in shaping how students interact with and evaluate LLM outputs, even under a single, consistently applied instructional design. This work contributes empirical evidence on how LLM adoption, role conceptualization, and verification behavior vary across computer science subfields, extending our prior work on structured, reflective LLM instruction to show how its effects differ by task rather than only in aggregate.
Read the full abstract ▾
All authors Sehrish Basir Nizamani, Yoonje Lee, Nikitha Donekal Chandrashekar, Margaret Ellis, Naren Ramakrishnan
Probability
Wetting Transition of the Two-Dimensional Solid-on-Solid Model with Quenched Disorder
Seokun Choi
We study the wetting transition for a two-dimensional Solid-On-Solid (SOS) interface above a hard wall in the presence of quenched disorder. The interface is represented by a nonnegative integer-valued height function $\phi$, with
Hamiltonian $\mathcal H^\omega(\phi)=\beta\sum_{x\sim y}|\phi(x)-\phi(y)|-\sum_x(h+\alpha\omega_x-\lambda(\alpha))\mathbf 1_{\{\phi(x)=0\}}$, where $(\omega_x)_{x\in\mathbb Z^2}$ is an i.i.d. centered field, $\alpha\ge0$ is the disorder strength, and $\lambda(\alpha)=\log\mathbb E[e^{\alpha\omega_0}]$. This normalization is chosen so that the annealed model coincides with the corresponding homogeneous wetting model. As the wall attraction $h$ increases, the interface undergoes a transition from a delocalized phase, in which contacts with the wall have vanishing density, to a localized phase with a positive density of contacts. For sufficiently large $\beta$, the homogeneous wetting point and its sharp near-critical free-energy behavior are known. We prove that, for every fixed $\alpha\ge0$, the quenched critical point coincides with the homogeneous wetting point, $h_c(\beta,\alpha)=h_w(\beta)=-\log(1-e^{-4\beta})$. Moreover, writing $u=h-h_w(\beta)$, we show that the quenched excess free energy satisfies $\overline F_\beta(\alpha,u)=\overline F_{\mathrm{homo}}(\beta,u)+o(u^3)$ as $u\downarrow0$, where $\overline F_{\mathrm{homo}}(\beta,u)$ denotes the leading homogeneous wetting asymptotic and satisfies $\overline F_{\mathrm{homo}}(\beta,u)\asymp u^3$. Thus, at the level of both the critical point and the leading near-critical free-energy asymptotics, the quenched model has the same behavior as the homogeneous SOS wetting model. This contrasts with the disordered SOS pinning model, in which quenched disorder leaves the critical point unchanged but modifies the leading critical behavior of the free energy.
Read the full abstract ▾
Combinatorics
A Proof of the Third and Cubic Borwein Conjectures
Yicen Ma
We establish the coefficient sign patterns in the Third Borwein
conjecture and the modulus-three Cubic Borwein
conjecture. The analytic arguments apply for $n\ge1750$ and $n\ge500$, respectively. They combine exact dissections and positive coefficient identities near the boundary with saddle point estimates that preserve cancellation between primitive-root contributions. Paired Gaussian estimates remove the leading odd error, and explicit remainder bounds cover the complementary contours. The remaining finite intervals are checked by exact integer arithmetic. The Third finite verification through $1749$ is author-confirmed; the Cubic verification through $500$ is supported by two complete integer implementations and additional coefficient crosschecks.
Read the full abstract ▾
Representation Theory
Homology and higher-order intermediate categories for proper abelian subcategories
Zheng-Gang Wu and Xianhui Fu
Without needing a $t$-structure, we establish natural long exact homology sequences on finite extension windows with $\ell+1$ layers of extension-closed proper abelian subcategories, assuming vanishing of negative self-extensions in degrees $1,\ldots,\ell+1$, and prove uniform sharpness of this bound. In the Krull-Schmidt setting, we classify intermediate subcategories stable under taking pure homology layers by compatible sequences.
Read the full abstract ▾
High Energy Physics (Phenomenology)
Analyticity Bootstrap of Feynman Integrals
Xiang Li, Dao-Ming Mu and Yan-Qing Ma
We propose an analyticity bootstrap method to determine the reduction of Feynman integrals (FIs). The key observation is that, at any kinematic point, a dimensionally regularized FI possesses a Taylor expansion region (analytic region), and the finiteness of the Taylor region strongly constrains the integral reduction coefficients, which are rational functions of the kinematic variables. By writing the most general ansatz compatible with the singularity structure and then fixing the remaining parameters with only a few integration-by-parts (IBP) samples or asymptotic expansions, one can obtain the full reduction. We develop the method systematically in both single-variable and multi-variable cases, with many explicit examples. Interestingly, the differential equations (DEs) of the master integrals determine all these analyticity constraints and, conversely, the constraints facilitate the construction of the DEs. In particular, we show that the DEs in the auxiliary mass flow method can be constructed at significantly reduced cost and can even be fully determined without using IBP.
Read the full abstract ▾
Machine Learning
Looping Beyond Twice: A Scalable Recipe for Looped Mixture-of-Experts
Di He, Pengxiang Li, Da Chang et al. (+3 more)
Looped
Transformers introduce recurrent depth as a new scaling axis for LLMs: by repeatedly applying shared
Transformer blocks, they increase effective depth without increasing parameter count. However, the benefits of looping remain unclear for large MoE LLMs under FLOPs-matched comparisons. The main reason is that the gains from additional iterations diminish quickly and can even turn into degradation, so the extra FLOPs spent on looping yield little substantial improvement. Consequently, prior work typically settles on two loops. We identify two main obstacles to scaling looped MoE. First, looping inherits and amplifies the curse of depth: hidden-state variance grows with each iteration as residual updates accumulate, which destabilizes deep recurrence and causes representations to drift. Second, looped MoE suffers from expert selection collapse: routers repeatedly select the same experts across loops, so extra iterations add computation without adding computational diversity. Guided by this diagnosis, we propose LOOM, built on a single principle: each loop should contribute new computation while keeping the recurrent state stable. LOOM stabilizes recurrence by scaling residual updates to bound variance growth and re-injecting the input
embedding at every loop, and diversifies it through per-loop routers that engage different experts and a Looping Residual that carries earlier outputs forward. Experiments across 100M-1.7B models show stable scaling to 9-12 loops. Under near-iso-FLOP, the 700M model performs best at 5 loops, reducing perplexity from 18.36 to 16.54 and improving average zero-shot accuracy from 38.84% to 39.53% over the non-looped baseline. Without FLOP matching, the 1.7B model trained on 60B
tokens peaks at 9 loops, reducing perplexity from 9.62 to 7.77 and improving average zero-shot accuracy from 42.4% to 47.7%. Code is available https://github.com/hed-ucas/LOOM.
Read the full abstract ▾
All authors Di He, Pengxiang Li, Da Chang, Qingyan Meng, Lu Yin, Shiwei Liu
High Energy Physics (Theory)
The Love for Three Logarithms: Universal Running of Dynamical Love Numbers
Sebastian Garcia-Saenz
The static
tidal Love numbers of black holes vanish in general relativity (GR), but are generically non-zero, and in fact 'run' logarithmically with scale, in more general theories or beyond the static limit. We expose a model-agnostic framework for the running of Love numbers at the dynamical level, valid for any
tidal perturbation on any static, spherically symmetric, asymptotically flat background, based on the near-zone expansion in powers of the frequency $\omega$ and on the monodromy at spatial infinity. A master recurrence formula determines the monodromy generator to all orders in $\omega^2$, whose three independent entries fully encode the running of the Love numbers. This structure explains the quadratic nature of the renormalization group equation, here proved to all orders and for any type of Love number. We discuss the scheme dependence of the logarithmic coefficients, which is made transparent by the monodromy form. In particular, this clarifies that the anomalous exponents of the
tidal field, the so-called 'renormalized angular momentum' of the Mano-Suzuki-Takasugi formalism, are scheme-invariant and directly computable on any spacetime, without need of explicit solutions. We discuss several applications to black holes, including the derivation of a closed-form formula for the leading dynamical running of axial Love numbers of Reissner-Nordström black holes, proving in particular that it vanishes in the extremal limit. For generic black holes beyond GR, we prove that dynamical running probes deformations of the metric at lower multipoles than the static one, and we illustrate the power of the method by deriving some exact results for the Hayward and Bardeen regular black holes.
Read the full abstract ▾
Statistical Methodology
Spatio-temporal coarse-grained Hawkes processes
Tomoya Uda and Shinsuke Koyama
Spatio-temporal Hawkes processes are widely used to model self-exciting event data, but conventional inference methods require access to the occurrence time and location of every event. In many applications, however, observations are available only as counts aggregated over temporal intervals and spatial regions. We propose a spatio-temporal coarse-grained Hawkes process, a discrete-time count model that approximates aggregated observations generated by an underlying spatio-temporal Hawkes process. The proposed model constructs effective excitation kernels by averaging the temporal and spatial triggering effects over spatio-temporal bins while explicitly incorporating excitation occurring within the same temporal bin. This formulation enables direct modeling of aggregated count data without introducing latent event times or locations. We characterize the first- and second-order properties of the proposed process and derive asymptotic approximation errors for the stationary mean and autocovariance relative to those of the corresponding aggregated Hawkes process in the joint limit of fine temporal and spatial discretization. The analysis shows that the proposed model provides a higher-order approximation than the conventional binned Poisson approximation by accounting for intra-bin excitation. Numerical experiments confirm the theoretical results and demonstrate accurate approximation over a broad range of aggregation scales. We further develop a moment-based estimation procedure and apply the proposed framework to earthquake occurrence data through an aggregated Epidemic-Type Aftershock Sequence (ETAS) model. The results indicate that the proposed approach provides a computationally efficient alternative for inference and prediction from spatio-temporally aggregated event data.
Read the full abstract ▾
Computation and Language
BanglaDial-Abuse: A Corpus-Grounded Dataset for Regional Dialect Identification in Abusive Bangla Text
Hasin Almas Sifat
Regional linguistic variation remains an important challenge for Bangla natural
language processing, particularly in informal and non-standard text. This paper introduces BanglaDial-Abuse, a balanced Bengali-script
dataset developed for regional dialect identification in abusive and hostile Bangla text. The
dataset contains 1,000 sentences distributed equally across four linguistic varieties: Standard Bangla, Chattagram, Sylhet, and Barishal, with 250 samples per class. The resource was constructed using a corpus-grounded synthetic procedure incorporating regional variation in pronouns, possessive forms, verb morphology, negation, interrogative structures, postpositions, vocabulary, and Bengali-script spelling conventions while preserving the underlying hostile or abusive meaning. Descriptive analysis shows broadly comparable sentence-length distributions but partially distinct lexical spaces across the four classes. Pairwise Jaccard vocabulary similarity ranges from 0.37 to 0.56. The primary task is four-class regional dialect identification rather than binary abusive-text detection. The
dataset is publicly available through Zenodo under a Creative Commons Attribution 4.0 license. The current version is intended as a research and prototyping corpus rather than a native-speaker-validated gold-standard linguistic resource. Keywords: Bangla, Bengali, dialect identification, regional dialect, abusive
language, low-resource NLP, Chattagram, Sylhet, Barishal,
dataset
Read the full abstract ▾
Data Structures and Algorithms
When Is Deletion Ordering Tractable? From Update Dynamics to Permutation Structure
Xinyu Wang, Ziyu Zhao, Yixuan He et al. (+1 more)
Given a fixed set of pending deletion requests, retraining from scratch after each request is prohibitive, so a prescribed request-wise policy processes them sequentially. The resulting terminal model can depend on their order. Rather than prescribing an ordering rule, we study the permutation objective induced by the fixed policy and ask when it admits simpler structure. We identify two independent reductions: position additivity represents the objective by request–position costs, reducing optimization to assignment and, with a shared positional profile, sorting; suffix localization removes dependence on the distant prefix while retaining interactions among the surviving requests. Under shared affine updates, we characterize the quadratic interactions that obstruct additivity, prove the reductions' independence, and show that suffix-conditioned assignment improves the approximation rate from O(p^L) toO(p^(2L)). Experiments recover both structures in executed objectives. A controlled damped-Newton sweep shows that stronger contraction shifts the objective toward shorter, more suffix-specific dependence, while two full-network policies exhibit distinct positional and within-suffix structure. Structures identified from compact execution sets also predict unseen orders. These results frame deletion ordering as identifying the computational structure induced by the executed updates.
Read the full abstract ▾
All authors Xinyu Wang, Ziyu Zhao, Yixuan He, Xiaowen Chang Alex Smola
Computer Vision
OptimusMesh: Compact Autoregressive Mesh Generation from Point Clouds via Sparse Latent Pivots
Mazhar Iqbal, Xuanmeng Sha, Naoya Chiba et al. (+2 more)
Generating compact and geometrically faithful 3D meshes directly from point clouds remains a fundamental challenge. Point clouds are unordered and sparse, whereas meshes exhibit irregular structure and varying
topology. As a result, many existing approaches rely on implicit representations followed by surface extraction or reconstruction. Although effective, these pipelines can produce dense or over-smoothed meshes, often requiring computationally expensive post-processing and simplification. We present OptimusMesh, a framework for direct compact triangle mesh generation from point clouds using sparse latent pivot conditioning. Our key idea is to compress $2{,}048$ oriented input points into only $16$ sparse latent pivots, reducing the
geometric conditioning set by $128\times$. These pivots provide a compact structural representation shared across a two-stage autoregressive framework that first generates mesh vertices and then predicts triangular faces conditioned on the generated vertices and the same pivots. Compared with the evaluated recent point-cloud-conditioned autoregressive methods, which use $257$ decoder-conditioning
tokens, OptimusMesh uses only $16$, yielding a $16.1\times$ shorter conditioning sequence. Experiments show that OptimusMesh produces the most compact outputs among the compared recent autoregressive methods, using $25.7\%$–$94.1\%$ fewer faces while maintaining competitive
geometric fidelity and distributional quality.
Read the full abstract ▾
All authors Mazhar Iqbal, Xuanmeng Sha, Naoya Chiba, Yuki Uranishi, Tomohiro Mashita
Computers and Society
Virtual Global Collaboration in Data Analytics and Machine Learning Education: A Mixed-Methods Study of Asynchronous Cross-Border Teamwork
Sehrish Basir Nizamani, Saad Nizamani, Khyati Goyal et al. (+2 more)
This innovative practice full paper examines a structured Virtual Global Collaboration (VGC) activity between undergraduate computing courses in the United States and Pakistan. The project was designed to support technical and intercultural skill development through asynchronous international teamwork in data analytics and introductory machine learning. Students worked in mixed-institution teams through a six-phase project including cultural orientation,
dataset selection, data cleaning, analysis, introductory machine learning, and structured reporting. Using shared computational tools, teams coordinated across time zones to complete a joint data-driven project. Survey and qualitative reflection data were analyzed to examine collaborative experience, communication and cultural dynamics, learning outcomes, perceived value, and global readiness. Results show consistently positive student experiences across both cohorts, with collaborative processes strongly associated with learning and perceived value, and cross-cultural communication emerging as the primary driver of global readiness. These findings demonstrate how short-term, structured VGC can be effectively integrated into computing courses to support both technical learning and global competence.
Read the full abstract ▾
All authors Sehrish Basir Nizamani, Saad Nizamani, Khyati Goyal, Sarwat Nizamani, Zannah Zeiw
Algebraic Geometry
Verdier-Type Category, Geometric Serre's conditions and perversity
Shang Xu
We generalize Serre's conditions for coherent sheaves to triangulated categories equipped with a $t$-structure and duality, which we call Verdier-type categories. We present examples of such categories and their associated Serre conditions, then focus on constructible sheaves, where we refer to these conditions as
geometric Serre conditions. We investigate their relationship with perversity and illustrate the theory through a range of examples. Finally, we prove a
geometric version of the Auslander-Reiten
conjecture.
Read the full abstract ▾
Chemical Physics
Toward Long-Range Correlation Energies from Charge Fluctuations: Incorporating Exact Exchange into ACKS2$\omega$
YingXing Cheng
Long-range electronic correlation can be represented in terms of coupled fluctuations of atomic charges and higher multipoles, motivating a density-response description beyond local dipolar polarization, particularly in low-dimensional, conducting, and small-gap systems where collective electronic fluctuations become important. We extend the frequency-dependent polarizable force field ACKS2$\omega$ to electronic-structure references containing exact exchange by deriving both the static and additional frequency-dependent exchange contributions, with particular
attention to HFsrPBE, which combines long-range Hartree–Fock exchange with short-range PBE exchange and correlation. Numerical validation of ACKS2$\omega$ using dipole polarizabilities and $C_6$ dispersion coefficients for the 27 molecules of the TS27 set, with HF, PBE, B3LYP, and HFsrPBE as the underlying electronic-structure references, shows that including the exchange contributions and enriching the response basis improve agreement with the corresponding orbital-space linear-response calculations, although finite-response-basis errors remain molecule dependent. Long-range correlation energies evaluated through the adiabatic-connection fluctuation-dissipation
theorem using the ACKS2$\omega$ parameters reproduce the corresponding HFsrPBE orbital-space results for small test molecules when the response basis is made complete within the chosen orbital basis, while a local multipole decomposition resolves the finite-basis energies into monopole–monopole, mixed, dipole–dipole, and remaining higher-multipole contributions, providing a route to evaluating long-range correlation energies in terms of charge fluctuations within atom-condensed models.
Read the full abstract ▾
Computational Physics
An S-matrix Formalism for the Nonclassical Optical Response of Plasmonic Nanowires
Xin Zheng, Christos Tserkezis, Christos Mystilidis et al. (+2 more)
This paper reports a computational scheme which predicts the light-matter interaction in multiple mesoscopic plasmonic nanowires (NWs), each of which includes concentric or eccentric constituent wires. Lying at the core of the proposed scheme is the S-
Matrix of each individual cylindrical interface. The optical response of the materials inside and outside each interface can be described by three popular mesoscopic models: the Hydrodynamic Drude model (HDM) and its diffusive variant, namely the Generalized Nonlocal Optical Response (GNOR) model, as well as the Surface Response Model (SRM). To take into account the relative displacements between the cylindrical interfaces, the addition
theorem is applied to construct the main equation behind the
algorithm, where the possibility of longitudinal waves intrinsic in the HDM and GNOR models is also included. The optical responses calculated by the proposed S
Matrix Method are compared with an in-house developed boundary element method (BEM) toolbox for a single gold NW and an eccentric dual-material NW, and excellent agreements are seen. The
algorithm is further physically checked for a sodium nanolens trimer featuring large field focusing, and the effects of the HDM and the SRM are compared.
Read the full abstract ▾
All authors Xin Zheng, Christos Tserkezis, Christos Mystilidis, Guy A. E. Vandenbosch, Xuezhi Zheng
Machine Learning
Parameter-Efficient Distributionally Robust Adaptation of Tabular Foundation Models under Subpopulation Shift
Seonghwi Kim, Sung Ho Jo and Minwoo Chae
Despite strong mean accuracy, tabular foundation models (TFMs) can perform poorly on underrepresented groups under subpopulation shift, where group proportions change between training and deployment. We propose DR-TFM, a parameter-efficient distributionally robust adaptation framework that requires no true group annotations. DR-TFM adjusts
attention to labeled context examples by fine-tuning an existing query scaling network or adding and training one, while keeping all other parameters fixed. We instantiate the framework with two robust objectives using estimated groups or source conditional distributions derived from training data. For TabPFN-3, adaptation updates only 0.016% of the pretrained model's parameters. Across five tabular
benchmarks, DR-TFM achieves substantially higher average worst-group accuracy than pretrained TFMs and the compared robust baselines without true group annotations, while maintaining competitive mean group accuracy. DR-TFM also improves average worst-group accuracy on ACS Income and across four additional TFMs.
Read the full abstract ▾
Molecular Networks
Protocol for Discovering Convergent Molecular Networks across Multi-Omics Datasets Using Network Propagation
Charles O. Olwal, Declan M. Winters, Immy A. Ashley et al. (+9 more)
Phenotypic outcomes arise from coordinated changes across molecular layers. Although multi-omics data capture complementary molecular signals, integrating them into coherent biological mechanisms remains challenging due to limited overlap at the gene level. Network propagation maps significant gene hits from each
dataset onto a prior knowledge molecular interaction network and diffuses their signals to neighboring genes. Mathematical integration of propagated networks, followed by clustering and pathway enrichment, reveals shared molecular networks, pinpointing the core biology captured across orthogonal molecular layers.
Read the full abstract ▾
All authors Charles O. Olwal, Declan M. Winters, Immy A. Ashley, Valerie Vandersall, Amberin Rahman, Alexandra Lindsey Djomkam Zune, Kyla F. Dangerfield, Yennifer Delgado, Sara K. Makanani, Sophie F. Blanc, Shipra Sharma, Mehdi Bouhaddou
Quantum Physics
Classical Hardness of Learning Functions of Hamiltonians
Sota Hashimoto and Akinori Kawachi
Morohoshi, Nakayama, Manabe, and Mitarai proposed a physically motivated
quantum machine learning problem in which the goal is to predict quantities of the form $\operatorname{Tr}[f(H)\rho]$ from classical descriptions of a
Hamiltonian $H$ and a
quantum state $\rho$, where $f$ is an unknown function. We call this problem
Hamiltonian function learning in this paper. They constructed an efficient
quantum learning
algorithm under suitable conditions, while leaving a rigorous proof of average-case classical hardness open. In this paper, we rigorously prove the average-case classical hardness for two distribution-specific
Hamiltonian function learning problems for $f_{\cos,\pi}(\lambda)=\cos(\pi\lambda)$ and $f_{\exp,\beta}(\lambda)=e^{-\beta\lambda}$ discussed in the paper of Morohoshi et al. under the assumption of the average-case hardness of factoring random RSA
moduli. More specifically, we show that an efficient classical randomized learner under squared loss whose output hypotheses are evaluable in classical
polynomial time for either problem would yield a classical randomized
polynomial-time
algorithm for factoring random RSA
moduli.
Read the full abstract ▾
Artificial Intelligence
ReSolve: Reusing Candidate Reasoning through Selective Generative Moderation
Bangji Yang, Jiajun Fan, Hongbo Ma et al. (+6 more)
Sampling multiple solutions spends computation on intermediate deductions and unfinished arguments as well as final answers. We introduce ReSolve, a training-free inference procedure that reuses this candidate
reasoning through selective generative moderation. An answer-distribution controller invokes a model to examine existing derivations when candidates disagree or lack a parseable answer, then incorporates the generated solution into a bounded loop. Under Hybrid scoring on 130 competition-mathematics problems evaluated with two independently sampled candidate pools, ReSolve obtains 100 and 99 correct answers, compared with 91 and 92 for voting over the same four candidates, with no correct-to-incorrect changes relative to that vote in either pool. Eight-sample self-consistency obtains 94 and 96 correct answers while consuming substantially more
tokens; ReSolve uses 46.3% and 47.2% fewer
tokens in the two evaluations. A controlled ablation removes visible derivations while retaining answer keys, vote counts, and the per-state output-cap rule, reducing accuracy from 100 to 93 correct despite increasing computation. Selective and always-on Uniform moderation both solve 97 problems, while selectivity reduces moderation
tokens by approximately 54% and total pipeline
tokens by 6.2%. These results support candidate
reasoning as reusable inference computation. They do not establish an accuracy advantage over additional sampling or a distinct benefit from specialized route instructions.
Read the full abstract ▾
All authors Bangji Yang, Jiajun Fan, Hongbo Ma, Xi Zhu, Weizhi Zhang, Minghao Guo, Ye Li, Hamid Palangi, Jiaxuan You
Information Retrieval
Do Multilingual Encoders Produce Language-Consistent Semantic IDs?
Abhinav Bohra and Anuj Bohra
Semantic IDs (SIDs) compress item
embeddings into discrete code sequences used in generative
retrieval. We ask whether a
multilingual encoder is sufficient for different-
language renderings of the same product to receive
language-consistent SIDs. Using Amazon ESCI listings rendered in English, Spanish, and Japanese, we test whether translations remain close to their English source, whether residual quantization is unusually sensitive to translation-induced movement, and whether
multilingual or
language-balanced quantizer fitting improves SID agreement.
Multilingual E5 places translations measurably apart: under an English-heavy fit, a Japanese translation preserves the first SID code of its English counterpart in only 7.7% of cases, compared with 89.0% for an English rewording. Distance-matched product-directed controls produce nearly the same full-SID mismatch as translation, providing no evidence that the quantizer selectively amplifies
language directions. Balancing the fitting mixture makes codebook use more uniform but further reduces cross-lingual prefix agreement: Spanish first-code consistency falls from 28.3% to 6.6%, while an English-only fit preserves it for 67.6% of Spanish translations. These results show that
multilingual exposure and balanced codebook use alone do not guarantee
language-consistent SIDs.
Read the full abstract ▾
Artificial Intelligence
Auditing Action Settlement in LLM Agent Environments: Order, Progress, and Replay
Haotian Chen, Bowen Ye, Yuning Zhang et al. (+1 more)
Concurrent actions in large
language model (LLM)
agent environments require arbitration even when each proposal is individually valid. We implement a typed snapshot-settlement contract and audit three distinct properties: order sensitivity, useful progress, and replay consistency. Five settlement policies are tested in 28,800 exhaustive permutation trials and 2,160 scripted multistep episodes. Joint policies are spatially order-invariant conditional on fixed priorities, yet conservative rejection completes only 31.25% of
agents in a six-
agent doorway task versus 90.28% for random tickets; the paired improvement is 59.03 percentage points (95% bootstrap interval: 50.00-68.06). All policies preserve the tested spatial constraints, and priority arbitration still misses the independent small-instance optimum. A separate full-state journal audit exactly replays 156 checkpoints and rejects 1,332 constructed corruptions with a retained terminal anchor. The evidence concerns execution semantics, not human realism or long-run
fairness.
Read the full abstract ▾
All authors Haotian Chen, Bowen Ye, Yuning Zhang, Jingkun Yu
Combinatorics
Covering Bridgeless Claw-Free Cubic Graphs with Perfect Matchings
Huaibin Zhou and Xiumei Wang
The Berge
conjecture [Proc. London Math. Soc., 1979] asserts that every bridgeless cubic graph can be covered by five perfect matchings. The 5-cycle double cover
conjecture, proposed independently by Preissmann (1981) and Celmins (1984), asserts that every bridgeless graph admits five even subgraphs such that every edge belongs to exactly two of them. Hakobyan and Mkrtchyan [Ars Math. Contemp., 2019] proved that the 5-cycle double cover
conjecture holds if and only if every bridgeless claw-free cubic graph can be covered by four perfect matchings. In this paper, we prove that the Berge
conjecture holds for every bridgeless claw-free cubic graph $G$ and that $G$ has four perfect matchings covering at least $\frac{299}{315}|E(G)|$ edges.
Read the full abstract ▾
Astrophysics of Galaxies
Do Submillimeter Galaxies Trace Megaparsec Large-scale Structures? – An Overdensity Analysis of 449 Submillimeter Galaxies in COSMOS
Canze Deng, Daizhong Liu, Helmut Dannerbauer et al. (+19 more)
We present an environmental analysis and overdensity catalog for a large sample of submillimeter
galaxies (SMGs) in the COSMOS field in order to understand whether SMGs are signposts of megaparsec-scale overdensity structures. We apply the Poisson Probability Method (PPM) protocluster finder to characterize the significance ($\sigma$) and overdensity value ($\delta$) of any overdense structures surrounding specific 'beacon'
galaxies. These beacons include 449 SMGs from A3COSMOS (including 168 with spectroscopic
redshifts, 52 of which have CO or [CII] 158$\mu$m line detections), as well as a control sample of over 5000 main-sequence
star-forming
galaxies (SFGs) selected from the 40-band photometric-
redshift catalog in COSMOS2025. We successfully reproduce 15 known protoclusters and clusters in COSMOS (including Hyperion large-scale structure, CL J1001+0220 at z~2.5, and AzTEC-3 protocluster at z~5.3), and report 26 new PPM-based high-overdensity ($\delta$ $\geq$ 5), high-significance ($\sigma$ $\geq$ 5), robust candidates. We measure an overdensity fraction of 11.5-71.2% (1.3-61.0%) in the CO/[CII] 52 SMG (full 449 SMG) sample with different criterion categories, confirming that SMGs are signposts of overdensities compared to the control SFG sample. We also notice that the fraction is lower than 100%, indicating that not all SMGs are associated with overdensities. We thus inspect the JWST morphology and argue that the triggering of starburst in SMGs must have non-environmental mechanisms, for instance, merger and violent disk instability. Finally, we further conduct a spatial distribution analysis, finding that the location of SMGs in overdensities has a
redshift evolution, with SMGs being more concentrated in the inner regions of overdensities at higher
redshift.
Read the full abstract ▾
All authors Canze Deng, Daizhong Liu, Helmut Dannerbauer, Gianluca Castignani, Rosa Calvi, Lauro Moscardini, Benjamin Magnelli, Sylvia Adscheid, Eva Schinnerer, Hui Li, Rafael Arango-Toro, Caitlin Casey, Jacky S.W. Chan, Andreas Faisst, Maximilien Franco, Zohreh Ghaffari, Santosh Harish, Jeyhan Kartaltepe, Greta Toni, Jorge Zavala, Xin Zeng, Xianzhong Zheng
Computer Vision
The RSNA Intracranial Aneurysm (RSNA-ICA) Dataset
Maria Correia de Verdier, Rachit Saluja, Jason Sho et al. (+37 more)
Intracranial aneurysm rupture is associated with substantial morbidity and mortality, yet aneurysm detection remains challenging, particularly for small lesions and on routine non-angiographic imaging examinations. To support the development and evaluation of artificial intelligence (AI)
algorithms for intracranial aneurysm detection and localization, the Radiological Society of North America (RSNA), in collaboration with the American Society of Neuroradiology (ASNR), the Society of Neurointerventional Surgery (SNIS), and the European Society of Neuroradiology (ESNR), curated the RSNA Intracranial Aneurysm (RSNA-ICA)
Dataset. Developed for the 2025 RSNA Intracranial Aneurysm Detection Challenge, RSNA-ICA is a large, publicly available, expert-annotated
dataset comprising 7202 CTA, MRA, and MRI series from 4278 adult patients collected across 21 institutions in 12 countries spanning five continents. The
dataset includes 2566 CTA, 2166 MRA, and 2470 MRI series from patients with and without intracranial saccular aneurysms, providing substantial geographic and imaging diversity. Expert annotations indicate both aneurysm presence and location, and 178 series additionally include three-dimensional
segmentations of challenge-defined vascular locations. RSNA-ICA was used to develop and evaluate
algorithms in the 2025 RSNA Intracranial Aneurysm Detection Challenge. Of the 7202 image series, 5041 are publicly available through MIRA, while the remainder were used for challenge public and private test sets. The
dataset is freely available to the research community for noncommercial use and provides a comprehensive resource for advancing AI-based aneurysm detection across both angiographic and routine neuroimaging examinations.
Read the full abstract ▾
All authors Maria Correia de Verdier, Rachit Saluja, Jason Sho, Maryam Vabarizad, Rennie Yung-Chieh Chen, Uyen N. T. Nguyen, Mona Alrehaili, Layal Aweidah, Deniz Bulja, Wesley C. Chan, Hernan Chaves, Madhavi Duvvuri, Huseyin Ekin Ergin, Undrakh-Erdene Erdenebold, Ekim Gumeler, Mohamed Sobhi Jabal, Chin-Chi Kuo, Fatima Mubarak, Sevde Nur Emir, Scott Riley K. Ong, Johanna Ortiz, Almudena Pérez-Lara, Andreas M. Rauschecker, Shayan Sirat Maheen Anwar, Charit Tippareddy, Tam Tran, Sorawis Visrutaratna, John Mongan, Adam E. Flanders, Robyn Ball, Greg Zaharchuk, Peter D. Chang, Felipe Kitamura, Errol Colak, Luciano Prevedello, Tyler Richards, Data Contributor Group, Dataset Annotator Group, Evan Calabrese, Jeffrey D. Rudie
Computer Vision
Open Vocabulary Word Recognition From Transcribed Bangla Texts
Faias Satter and Sk. Md. Masudul Ahsan
An optical character recognition (OCR) can scan a paper and extract text using technology, making people's jobs easier. While various OCR systems are available in the software industry, finding a reliable equivalent solution for Bangla takes much work. When it comes to handwritten texts, the situation is much more unusual. Recognizing words from word images is the most critical stage in any OCR process. It is the second stage after segmenting words from text pictures. If this stage fails, the overall performance of the OCR will be poor, regardless of how well the other phases perform. This study aims to recognize words using deep learning in a handwritten Bangla word image. Three object detection models, SSD with MobileNetV2, Faster R-CNN with InceptionResNetV2, and an ensemble model of these two, have been used to train and test handwritten word images. A modified Non-Maximum Suppression has been introduced to enhance the effectiveness of the models' results. A customized
dataset of 9841 handwritten Bangla word images has been compiled, featuring diverse handwriting styles from various individuals. All three models' performances have been checked against the test
dataset, and the ensemble model has been the most impressive, with an F1-score of 92.61%. Also, at the word level, the ensemble model correctly recognizes 96.12% of the words to some extent. The system can be further improved by introducing a post-processing phase to correct errors generated by the system.
Read the full abstract ▾
Machine Learning
Does Scaling Reinforcement Learning Really Require More Training?
Bangji Yang, Jiajun Fan, Hongbo Ma et al. (+2 more)
Scaling
reasoning typically spends more compute on reinforcement learning (RL) or on inference. We show that a completed RL training history can yield policies stronger than the checkpoints visited by its optimizer. We call this policy-space scaling: expanding the deployable policy set accessible from a fixed RL history, without extending training or increasing per-query inference computation. We instantiate it with SURGE (Scaling Up RL Gradient-free via Eigenspace fusion). SURGE combines two checkpoints from the same RL run: a high-accuracy anchor and a competitive donor that generates shorter responses. It expresses both checkpoints as changes from their shared initialization, then spectrally decomposes the anchor's update to retain its dominant component and incorporate the donor's complementary component. With a fixed target for how much of the anchor update to retain, SURGE determines the block size from the weights without testing candidate policies. We evaluate two 1.5B mathematical-
reasoning histories, DeepSeek and Nemotron, and one 7B coding history, OLMo. SURGE improves
benchmark-average accuracy over both input checkpoints while using fewer
reasoning tokens than the anchor. It reaches 54.17% on DeepSeek AIME24 against a measured native maximum of 50.83%, and 83.7% on OLMo HumanEval+ against 82.8%. These gains exceed the observed training curves.
Geometric controls support the importance of RL-update structure beyond weight displacement or
token reduction alone. Each constructed model runs as a single policy. Our findings identify stored RL history as a reusable scaling resource: the capability available from a training run need not end at its best checkpoint.
Read the full abstract ▾
All authors Bangji Yang, Jiajun Fan, Hongbo Ma, Ruihan Guo, Ge Liu
Operator Algebras
Gromov-Hausdorff Convergence of Spectral Truncations for Noncommutative Tori
Vibhor Bhatt, Satyajit Guin and Palash Sharma
We prove the Gromov-Hausdorff convergence of the state spaces, equipped with Connes' distance function, of the operator systems associated with spectral truncations for noncommutative $n$-
tori, $n\ge 2$. The action of the $n$-
torus on the noncommutative $n$-
torus, together with an appropriate use of Bochner integrals, enables us to construct an $C^1$-approximate order isomorphism that establishes the convergence. The earlier convergence result for the classical case of
tori [20] follows as a special case of our result. We conclude by studying the $C^*$-envelopes and propagation numbers of the truncated operator systems and show that the structural properties of the truncated operator systems for
tori persist under deformation.
Read the full abstract ▾
Plasma Physics
Theory of Equivalent Tokamaks for Characterizing Turbulent Transport in Quasi-symmetric Stellarators
Hongxuan Zhu, R. Gaur, X. Wei et al. (+2 more)
It is well known that quasi-symmetric (QS) stellarators are isomorphic to tokamaks in terms of their neoclassical-transport properties, and the corresponding transport coefficients can be calculated in the same manner as in tokamaks. However, less is known regarding the turbulent-transport properties of QS stellarators, e.g. the transport coefficients from the ion-temperature-gradient (ITG) mode. In this work, a systematic theory of the “equivalent tokamaks” for QS stellarators is presented based on the local gyrokinetic formulation and the near-axis expansion theory. It is shown that to zeroth order in the minor radius, the equivalent tokamaks can be chosen to have circular flux surfaces and can be characterized by three
geometric quantities: the aspect ratio, the rotational transform, and the
magnetic shear. To achieve first-order accuracy, however, not all QS stellarators have equivalent tokamaks, but good approximations can be found for some cases either as global or local
equilibria. Local and global gyrokinetic simulations of ITG transport are performed for a selection of QS configurations, and quantitative agreement in the turbulent transport levels is found between the stellarators and their equivalent tokamaks.
Read the full abstract ▾
All authors Hongxuan Zhu, R. Gaur, X. Wei, Z. Lin, A. Bhattacharjee
Computational Geometry
Optimal Coresets for Hyperbolic Farthest-Point Queries via Ideal-Boundary Envelopes
Eunku Park
We study coresets for farthest-point queries in
hyperbolic space. Given a nonempty finite set $P \subset \mathbb{H}^D$ and $0<\varepsilon \le 1$, we seek a coreset $P_{\varepsilon} \subseteq P$ whose farthest distance from every query point underestimates that of $P$ by at most an additive $\varepsilon$ and retains at least a $1-\varepsilon$ fraction of it. For every fixed $D \ge 2$, we prove that the optimal worst-case coreset size is $\Theta\bigl(\varepsilon^{-(D-1)/2}\bigr)$. Our main
geometric ingredient is an exact reduction from
hyperbolic queries to an upper envelope on the ideal boundary. In the hyperboloid model, each input point induces a positive boundary-score function whose logarithm gives its asymptotic distance offset along geodesic rays. We define the ideal-boundary envelope as the pointwise maximum of these functions and prove that the supremum additive loss over all queries equals the maximum logarithmic gap between the input and coreset envelopes. For the upper bound, we move the minimum-enclosing-ball center to the origin and normalize the spatial coordinates, obtaining a bounded Euclidean point set whose boundary envelope is bounded away from zero. A standard Euclidean kernel then approximates all directional score maxima simultaneously, and the structure
theorem yields both guarantees. For the lower bound, a spherical packing on a fixed-radius
hyperbolic sphere, together with antipodal queries and the
hyperbolic cosine law, makes every input point indispensable, matching the upper bound even for either guarantee separately.
Read the full abstract ▾