Peer Revenue Impact factor: volatile

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Tuesday 22 September

Quantum Physics

Coherence-Enhanced Spatial Quantum Thermometry

Asghar Ullah, Giovanni Scala, Luis L. Sánchez-Soto et al. (+1 more)

Quantum thermometry has long promised sensitivities beyond classical limits, with recent paradigms highlighting quantum coherence as a primary driver of precision. While most studies focus on single-point temperature estimation, we extend this framework to the spatial domain. We consider a probe two-level system (TLS) coupled asymmetrically to an auxiliary TLS that equilibrates with a sample of profile T(x). The reduced probe state carries coherence in its local basis that remains temperature-sensitive after the probe transition has frozen out, yielding a Fisher information exponentially larger at low temperature than that of a population-only readout. An exact reparametrization converts the thermometric precision into a local spatial distinguishability bound; its validity requires only a sufficiently small relative temperature uncertainty and yields an explicit measurement budget. Adding auxiliary TLSs shifts the array's optimal working temperature upward and broadens the useful sensing window, with the achievable gain ultimately limited by the total available coupling rather than by an indefinite scaling law. Finally, we propose a microwave Mach-Zehnder interferometer that maps the temperature-dependent probe coherence onto an experimentally accessible quadrature observable.
Read the full abstract

All authors Asghar Ullah, Giovanni Scala, Luis L. Sánchez-Soto, Özgür E. Müstecapl{\i}oğlu

High Energy Astrophysical Phenomena

SN2025aico: An Interesting Case Of $^{56}$Ni Mixing, Ejecta Asymmetries, and Dust Formation in a Type IIb Supernova

Grace Davis, Kyle Medler, Chris Ashall et al. (+8 more)

Stripped-envelope supernovae provide a window into how massive stars lose their layers, mix radioactive material, produce dust, and explode asymmetrically. We present optical, near-infrared (NIR), and mid-infrared (MIR) observations of SN~2025aico, a Type~IIb supernova in LEDA~35384. Our spectroscopic sequence, spanning $+1$ to $+167$ d after explosion, follows its evolution from a photospheric phase exhibiting both hydrogen and helium features to prominent helium emission and, ultimately, nebular-phase ejecta. Using He I $1.083$ and $2.0581\,\mu$m transitions, we investigated the kinematics and geometry of the helium-rich material. Both transitions exhibit a three-phase, non-monotonic velocity evolution: an initial rapid decline, a subsequent increase, and an eventual plateau. We interpret this behavior as evidence for limited outward mixing of $^{56}$Ni, such that radioactive energy deposition reaches the outer helium-rich ejecta progressively as the ejecta expand, producing the non-thermal electrons responsible for helium excitation. After $\sim100$ d, both NIR He I transitions develop double-peaked emission profiles. Similar structure in the oxygen emission indicates ejecta asymmetry. Comparison with other supernovae suggests a tentative connection between this structure and explosion energy, potentially linked to the delay between core collapse and explosion. Serendipitous JWST observations at $+124.4$d reveal an infrared excess. Modeling favors warm ($\sim800$–$1500$K) carbon dust newly formed in the ejecta, together with cooler carbon or silicate dust likely associated with pre-existing circumstellar material. SN~2025aico demonstrates how continuous optical-to-MIR observations can connect progenitor evolution, explosion physics, ejecta geometry, and dust production.
Read the full abstract

All authors Grace Davis, Kyle Medler, Chris Ashall, Willem B. Hoogendam, Cameron Pfeffer, David O. Jones, Shunsaku Horiuchi, Tyco Mera, Benjamin Shappee, Charlotte Ward, Grace Showerman

Data Structures and Algorithms

Fast Geometric Spanners via Approximate Nearest Neighbor Search

Alexandr Andoni, Manuel Paez, Krish Singal et al. (+1 more)

We study the problem of constructing metric spanners in general metric spaces in subquadratic time when given blackbox access to a fast algorithm for batch approximate nearest neighbor search. In particular, we show the following results for any metric space $\mathsf{M} = ([n], \mathsf{d})$ with aspect ratio $\Delta$ admitting a $c$-approximate batch nearest neighbor search algorithm with runtime $\tau_{\mathsf{M}}(n)$, (1) There exists an algorithm that, for any $k \in \mathbb{N}$, constructs an $O(c k)$-distortion spanner with $\tilde{O}(kn^{1+1/2k} \log \Delta)$ edges and runs in time $\tilde{O}(\tau_{\mathsf{M}} \cdot k n^{1/k} \log \Delta)$. (2) Any algorithm that learns at most $o(n^{1+1/k}/k)$ pairwise distances by querying a distance oracle and a blackbox batch nearest neighbor search oracle necessarily incurs $\Omega(c k)$ distortion. Our results entail that (truly) sub-quadratic time algorithms for spanner construction is equivalent to subquadratic time BANN (up to constant-factor losses). As a further application, we use our fast spanner constructions to obtain a fast algorithm for approximating the Wasserstein distance $\mathsf{W}_q$, for all $q > 1$, over any metric space admitting an efficient batch approximate nearest neighbor search algorithm. Together with recent new efficient algorithms for approximate nearest neighbor search in $\ell_p$ spaces, for $p > 2$, our results entail the first subquadratic time algorithms for spanner construction (with the stated size-distortion tradeoff) and $\mathsf{W}_q$ distance approximation over these metric spaces.
Read the full abstract

All authors Alexandr Andoni, Manuel Paez, Krish Singal, Tian Zhang

Functional Analysis

Truncation bands, unitality, and homomorphisms in truncated Riesz spaces: Boulabiar's problems and beyond

Mohamed Habibi and Hamza Hafsi

We answer three questions from Boulabiar's survey on truncated Riesz spaces. First, the truncation band projection property coincides with the principal projection property, with no completeness or Archimedean hypothesis needed, and the band projection onto a truncation band is then given explicitly by the truncation itself. Second, a truncation on any Riesz space is unital exactly when its fixed-point set has a supremum. Hence bounded truncations on $KB$-spaces are unital, and a nonunital truncated Banach lattice with bounded truncation contains a closed sublattice copy of $c_0$ and is never an $AL$-space, while on the $AM$ side the smallest truncation unitization norm is always an $M$-norm, though the largest one need not be. Third, a truncation homomorphism whose domain truncation is Archimedean need not be a Riesz homomorphism, as an explicit counterexample built from $c_0$ and $c_0/c_{00}$ shows, but it is one modulo truncation infinitesimals whenever the codomain space is Archimedean.
Read the full abstract

Computers and Society

Asset-Class Specific Sustainability Disclosure: Lessons Learned from the EU MiCA Regulation

Ulrich Gallersdörfer, Lena Klaaßen, Christoph Kreiterling et al. (+1 more)

Sustainability disclosure in the European Union has so far focused on corporate reporting and sustainability information attached to traditional financial products. Crypto-assets, with consensus-mechanism-driven environmental externalities and fragmented issuer structures, largely fell through this disclosure architecture. The EU Markets in Crypto-Assets Regulation (MiCA) introduces the first EU-wide, asset-class-specific sustainability disclosure regime for crypto-assets by mandating standardised sustainability indicators for both issuers and crypto-asset service providers. Drawing on a policy and legal analysis of MiCA and its Level 2 measures, and on early implementation evidence from public registers and market practice, this paper shows how the policy approach shifted from debates about restricting energy-intensive consensus mechanisms to a transparency regime built on quantitative metrics, machine-readable reporting, and methodological alignment with the broader EU sustainable-finance framework. The paper also highlights practical frictions in implementation, including data gaps, responsibility allocation between issuers and intermediaries, and cross-border supervisory fragmentation. The overview and quantitative summary of the ESMA Interim MiCA Register in this paper may also provide valuable information to regulators and market participants in the context of the European Commission's 2026 targeted consultation on the review of MiCA, including its specific question on environmental and sustainability reporting.
Read the full abstract

All authors Ulrich Gallersdörfer, Lena Klaaßen, Christoph Kreiterling, Christian Stoll

Astrophysics of Galaxies

Radio Quasars in the Cosmic Dawn: A Spectroscopic Sample from the DESI Survey

Anniek J. Gloudemans, Jinyi Yang, David M. Alexander et al. (+55 more)

We present the largest statistical spectroscopic sample of radio quasars at 4.7<z<6.7 to date. These have been selected from a dedicated Dark Energy Survey Instrument (DESI) target program and provide a robust sample to study the influence of radio jets on early quasar evolution. Using low-frequency radio data from the LOFAR Two Metre Sky Survey Data Release 3 (LoTSS-DR3) at 144 MHz, we identify 83 high-z radio quasars, 26 of which are classified as radio-loud (5-15% of 646 confirmed high-z quasars, defined as $R = F_{5\text{GHz}}/F_{4400A} >10$). With radio luminosities of $\sim10^{25-28}$ W/Hz, the radio emission in all 83 quasars is expected to be jet-dominated rather than powered by star formation. Analysis of their rest-frame ultra-violet (UV) spectra reveals broadly similar composite spectral properties and CIV velocity shifts compared to radio non-detected quasars. The radio-detected and radio-loud quasars exhibit slightly smaller Ly$\alpha$ and CIV equivalent widths on average, however, we find no significant evidence ($\sim1.3\sigma$) for a higher fraction of weak-line quasars (WLQs; ~10+/-4% versus ~4+/-1% for radio non-detected quasars). We further find that the broad absorption line (BAL) quasars are predominantly radio-quiet, and no correlation is observed between absorption index and radio-loudness, suggesting that radio emission is not produced by the BAL outflows themselves and are physically distinct phenomena. Our results demonstrate that powerful radio jets were already well-established by z~5.
Read the full abstract

All authors Anniek J. Gloudemans, Jinyi Yang, David M. Alexander, Arjun Dey, Xiaohui Fan, Victoria A. Fawcett, Martin J. Hardcastle, Stephanie Juneau, George K. Miley, Leah K. Morabito, Adam D. Myers, Swayamtrupta Panda, Huub J. A. Röttgering, Timothy Shimwell, Daniel J. B. Smith, Feige Wang, J. Aguilar, S. Ahlen, D. Bianchi, D. Brooks, T. Claybaugh, A. Cuceu, A. de la Macorra, P. Doel, S. Ferraro, J. E. Forero-Romero, E. Gaztañaga, S. Gontcho A Gontcho, G. Gutierrez, C. Hahn, M. Ishak, R. Joyce, R. Kehoe, D. Kirkby, T. Kisner, A. Kremin, M. Landriau, L. Le Guillou, M. E. Levi, M. Manera, A. Meisner, R. Miquel, J. Moustakas, S. Nadathur, W. J. Percival, F. Prada, I. Pérez-Ràfols, A. Robertson, G. Rossi, E. Sanchez, E. F. Schlafly, D. Schlegel, J. Silber, D. Sprayberry, G. Tarlé, B. A. Weaver, R. Zhou, H. Zou

Operator Algebras

A separably representable counterexample to Naimark's problem in ZFC

Ryotaro Tanaka

We construct in ZFC a unital, simple, infinite-dimensional $C^*$-algebra whose nonzero irreducible representations form a single unitary equivalence class, but which admits a faithful representation on a separable Hilbert space. The algebra contains a unital copy of the canonical anticommutation relation (CAR) algebra, and the normalized CAR trace has a unique extension among all states. This extension is tracial, and its GNS representation is separable and faithful, with weak closure the hyperfinite $\mathrm{II}_1$ factor. In contrast, every nonzero irreducible representation acts on a Hilbert space of density $2^{\aleph_0}$. The construction separates the added unitaries into shell terms, controlled by finite CAR relations, and rank-one defect terms. These relations determine every irreducible representation of the generated algebra and every extension of the CAR trace. The algebra has norm density $2^{\aleph_0}$; consequently, the continuum hypothesis (CH) is equivalent over ZFC to the existence of a counterexample to Naimark's problem of norm density $\aleph_1$.
Read the full abstract

Artificial Intelligence

Which Objectives Need a Dial? Predicting Objective Conflict and Covering Trade-offs in Steerable Pluralistic Alignment

David Tsoi and Esra Dönmez

People hold diverse, sometimes conflicting values, so no single aligned model can satisfy everyone. Pluralistic alignment therefore calls for steerable models that can balance competing objectives differently. Multi-Objective Direct Preference Optimization (MODPO) does this by using an objective weight to span a continuum of trade-offs. We study two questions: when can one model improve two objectives simultaneously, and how can many trade-offs be covered without training a separate model for each? Across seven objective pairs from HelpSteer and UltraFeedback, two pre-training measurements predict whether objectives align or conflict for human-annotated data, but not for AI-annotated data, where response length and repetition confound reward-model scores. For broader trade-off coverage, selecting the nearest trained model and merging model parameters both help, but neither consistently matches direct training. These findings yield practical guidance for building steerable models that serve diverse preferences.
Read the full abstract

High Energy Physics (Phenomenology)

Impact of neutral fluxes and signal significance optimization on semi-exclusive $pp \to t\bar{t}$ production via deep learning training

A. Cota Rodríguez, Jesús Alberto V. Corral and J. A. Murillo Quijada

Photon flux benchmark models and recent experimental estimates of pomeron energy fluxes and structure functions are implemented in Monte Carlo simulations to assess their impact on physical observables and signal strength uncertainty for the yet-to-be-observed semi-exclusive $t\bar{t}$ production in proton-proton ($pp$) collisions at Large Hadron Collider (LHC) and Future Circular Collider (FCC) energies. The expected cross sections increase by factors of approximately 50 and 22 for pomeron-induced and photon-induced processes, respectively, from LHC to FCC energies. TensorFlow deep neural networks discriminate semi-exclusive processes from the non-peripheral $t\bar{t}X$ background, achieving a test area under the receiver operating characteristic curve (AUC) of $0.9917 \pm 0.0002$ for the photon-induced signal. At detector level, using the CMS detector geometry at CERN, the minimum Hadronic Forward energy observable is identified as the most effective discriminator against this background. For low-pileup $pp$ data, an Asimov dataset analysis incorporating statistical uncertainties and systematic contributions from pomeron and photon schemes predicts a significance of $5\sigma$ for pomeron-induced and photon-induced $t\bar{t}$ production with integrated luminosities of 1 and 4.7 $\mathrm{fb}^{-1}$, respectively. The systematic contributions to the total uncertainty are 25.1% and 7.6%, respectively. These results highlight the potential for experimental observation using Run 2 and Run 3 LHC data and for further studies within the LHC forward physics program.
Read the full abstract

Artificial Intelligence

Building Socio-Affective Artificial Intelligence for Interactive Multi-Agent Simulations

David Berga

The objective of this article is to provide design principles and a software architecture for enabling interaction between humans and multiple agents in simulated dynamic worlds. This connects the current era of general artificial intelligence (AI/AGI) with the proliferation of transformer-based conversational agents and the increased computational capabilities. Given an overview of current and previous multi-agent theories of mind (socially and affectively-aware agents), the existence of an integrative design of agent interactions with themselves and with humans must be crucial for understanding how to create sustainable and governance in future human-agent reasoning systems. In this work is presented a software "AGIMUD" that integrates: A. socially-aware reasoning and emotion in agent behavior and interaction, B. a design of human multimodal scheme for human users, artificial agents and simulated worlds, and C. distributing the AI processing through the network to enable multiple autonomous agents. These integrations allow the dynamic world recreation as multi-user dungeons (MUDs) where both agents and humans can interact simultaneously in real time. Find the code online in https://github.com/dberga/AGIMUD.
Read the full abstract

Computation and Language

Experts Rise Where LLMs Disagree: Using Cross-Model Disagreement to Target Expert Effort in LLM Codebook Revision for Large-Scale Annotation

Zeyu He, Zhuqian Zhou, Kirk Vanacore et al. (+2 more)

Large-scale text annotation brings expert insight to millions of documents, often through a codebook that AI annotators follow. Developing a robust codebook, however, takes months. Large language models (LLMs) could speed this process by applying an early codebook to the data, surfacing cases with strong LLM disagreement, and eliciting expert feedback to address them. We examined three ways experts can provide feedback for LLM codebook revision: (i) editing LLM-generated revisions driven by cross-LLM disagreement (Codebook Verifying), (ii) answering questions about LLM disagreements (Question Answering), and (iii) labeling disagreement cases with rationales (Rationale Labeling). Experiments on thousands of tutoring-session transcripts show that Rationale Labeling yielded the highest LLM-labeling accuracy (64.9%) against expert labels, outperforming the expert-revised codebook (57.8%). The best Question Answering setting also outperformed it (60.5%). Our work shows that LLMs can be used to strategically target expert attention, shortening months of codebook revision to days without sacrificing labeling performance.
Read the full abstract

All authors Zeyu He, Zhuqian Zhou, Kirk Vanacore, Rene F. Kizilcec, Ting-Hao 'Kenneth' Huang

Statistical Mechanics

First Passage in the Presence of Jump Rate Fluctuations: A Transition from Infinite to Finite Mean

Bara Levit and Shlomi Reuveni

A simple symmetric random walk on a one-dimensional lattice is guaranteed to reach any target site, yet its mean first-passage time diverges. Here, we show that stochastic fluctuations in the jump rate can fundamentally alter this classic result. We find a phase transition in the mean first-passage time: sufficiently broad rate fluctuations render it finite, whereas weaker fluctuations leave it infinite. Our results demonstrate that temporal fluctuations in the jump rate—and hence in the diffusivity—can overcome spatial wandering and regularize otherwise divergent first-passage times.
Read the full abstract

Computer Vision

nnFoundation: 3D Foundation Models for Radiology

Constantin Ulrich Harsy, Tassilo Wald, Karol Gotkowski et al. (+80 more)

Radiological artificial intelligence has advanced rapidly, yet most systems remain narrowly task-specific, data-intensive, and fragile under domain shift. Foundation models promise more transferable and data-efficient solutions, but existing approaches are limited in scale, evaluated narrowly, and often assume that a single pretrained model can support diverse downstream tasks. Here we present nnFoundation, complementary convolutional and transformer-based 3D radiological foundation models. Developed within the Human Radiome Project (THRP), nnFoundation is trained on 2.1 million CT, MRI, and PET image volumes from 125 institutional and public datasets. We evaluate them across 108 tasks spanning segmentation, detection, classification, report generation, and image retrieval, including evaluations under domain shift, by external partners and in low-data and low-compute regimes. Across all task types, our convolution- and transformer-based nnFoundation models consistently outperform both prior 3D foundation models and training from scratch, establishing state-of-the-art performance for radiological imaging. However, performance follows a consistent task-dependent structure: the convolutional nnFoundation model dominates spatially localized tasks, whereas the transformer-based nnFoundation model excels in tasks requiring global semantic reasoning and in frozen-feature settings. Dynamically aligning the foundation model topology with the dataset characteristics post-hoc further improves transfer across heterogeneous 3D settings. These results show that transferable 3D radiological performance is governed not by a single universal model, but by the interplay of scalable pretraining, complementary architectures, and dataset-aware adaptation. We release nnFoundation models integrated into nnU-Net and nnDetection, enabling immediate application across established radiology workflows.
Read the full abstract

All authors Constantin Ulrich Harsy, Tassilo Wald, Karol Gotkowski, Yannick Kirchhoff, Marcel Knopp, Maximilian Rokuss, Elisa Stegmeier, Philipp Schader, Dasha Trofimova, Raphael Stock, Kim-Celine Kahl, Stephen Schaumann, Selen Erkan, David Zimmerer, Stefan Denner, Moritz Langenberg, Sebastian Ziegler, Katharina Eckstein, Maximilian Fischer, Jonathan Suprijadi, Bálint Kovács, Benjamin Hamm, Anand Deshpande, Dimitrios Bounias, Nico Disch, Shuhan Xiao, Jessica Kächele, Jan Sellner, Rajesh Baidya, Jeremias Traub, Lars Krämer, Maximilian Zenk, Tim Rädsch, Stefan Dvoretskii, Robin Peretzke, Jonathan Deissler, Alexandra Ertl, Partha Ghosh, Kris Dreher, Stefan Dinkelacker, Annika Reinke, Evangelia Christodoulou, Numan Saeed, Yoland Savriama, Santiago Estrada, David Kügler, Laura Alexandra Daza Barragan, Cristina Isabel Gonzalez Osorio, Jan Peeken, Michael Baumgartner, Marvin Teichmann, Guillaume Chabin, Matthias Kirchler, Valentin Koch, for the ALFA study, Markus Hohenhaus, Dimitri Koslov, Nina Decker, Mohammad Yaqub, Arnd Heuser, Martin Reuter, Julia A. Schnabel, Tobias Heimann, Florin Ghesu, Paul Brachmann, Claus P. Heußel, Alexander Radbruch, Gianluca Brugnara, Aditya Rastogi, Martha Foltyn-Dumitru, Heinz-Peter Schlemmer, Ignaz Reicht, Julius C. Holzschuh, Michael Bach, Bram Stieltjes, Kai Schlamp, Lena Maier-Hein, Marco Nolden, Ralf Floca, Paul F. Jäger, Philipp Vollmuth, Fabian Isensee, Klaus H. Maier-Hein

Computer Vision

A 3D Pose-Based Ensemble Framework for Cricket Shot Classification and Automated Biomechanical Analysis

Sourav Shome, M.D. Ashiquzzaman Rahad and Rameswar Debnath

Cricket is one of the most celebrated sports world-wide, and technological advancement has become deeply embedded in how the modern game is analyzed and coached. Cricket shot classification and automated performance analysis add a further dimension to this trend. Traditional approaches rely on RGB video features or static images, which are sensitive to environmental variations such as camera angle, lighting, and background clutter, and often fail to capture the underlying biomechanics of batting actions. In this paper, we propose a system to improve cricket coaching that takes raw video data, extracts batsmen from video frames using YOLO, and extracts 3D pose data from video frames using MeTRAbs. The system produces sequential skeletal pose data of 30 body points and captures the biomechanical features of a batsman. As part of the system, we also propose a deep learning ensemble for shot classification of four shots: flick, pull, defense, and drive. The ensemble performed well, compared to existing classification works, achieving 97.68% accuracy. In addition, we analyzed the misclassification rates to identify cases where shots were incorrectly classified and examined their possible causes. Our proposed system allows novice players to obtain useful feedback, such as important joint angles relative to expert batsmen, which can also be useful for injury prevention. The shot classifier also helps track class-wise shots over time for further analysis. In addition to novice players, coaches can use the system for player evaluation.
Read the full abstract

Algebraic Geometry

Euler characteristics of strata of $k$-differentials

Matteo Costantini, Alessandro Giacchetto, Martin Möller et al. (+2 more)

We compute the orbifold Euler characteristics of strata of $k$-differentials on the moduli space of smooth pointed curves. Our main result is a closed coefficient-extraction formula in terms of hyperbolic functions, valid for arbitrary strata and generalizing the Harer–Zagier formula for the Euler characteristic of the moduli space of curves. In particular, these Euler characteristics are polynomial in the prescribed orders at the marked points. The proof first uses Riemann–Roch and Serre duality to reduce the computation to the Hurwitz case $k=0$, which is then evaluated using the representation theory of the symmetric group and Fock-space techniques. We also give an intersection-theoretic expression for the same invariants as tautological integrals over double ramification cycles and derive a recursion which uniquely determines them from genus zero. For odd $k$ and odd orders, we introduce a spin-weighted refinement and formulate corresponding closed and intersection-theoretic conjectural expressions. Finally, we determine the large-genus asymptotics of the Euler characteristics.
Read the full abstract

All authors Matteo Costantini, Alessandro Giacchetto, Martin Möller, Adrien Sauvaget, Johannes Schmitt

Information Retrieval

Distilling Lexical Product Associations into Deep Transformers: An Extreme Multi-Label Approach for Natural Language E-Commerce Search

Sunnidhya Roy and Samarpita Bhaumik

Traditional e-commerce search platforms rely heavily on inverted indices and token-level lexical matching algorithms (e.g., BM25 and TF-IDF), which frequently fail on conversational, intent-driven, or paraphrased user queries – the classic vocabulary mismatch problem. We formulate conversational product recommendation as an Extreme Multi-Label Classification (XMLC) problem over an e-commerce catalog of N = 54,000 products spanning 27 balanced retail categories from the Amazon Reviews '23 benchmark. Using a pre-trained DistilBERT transformer encoder, we distill dense item-to-item similarity topologies (generated via TF-IDF cosine similarity over cumulative metadata with K = 50 nearest neighbours) into a deep contextual representation via a pseudo-label knowledge distillation framework. Evaluated on an exact 85/15 train/validation split (8,089 held-out products across C = 53,923 output classes) with strict self-exclusion enforced, the DistilBERT neural student achieves P@1 = 93.15%, P@5 = 90.08%, NDCG@10 = 0.8845, and MRR@10 = 0.9545, closely recovering the empirical ceiling established by the corrected TF-IDF teacher (P@1 = 98.10%, NDCG@10 = 0.9419, MRR@10 = 0.9882). Furthermore, a qualitative benchmark across ten structured natural language query archetypes – encompassing situational, cross-category, paraphrased, and negative-constraint queries – demonstrates that the transformer student generalises substantially beyond keyword matching, successfully resolving implicit user intent where lexical models fail completely. Finally, we analyse the architectural and memory scalability trade-offs of extreme classification projection layers at industrial catalog scale (> 10^6 items) and present a concrete deployment trajectory toward Dual-Encoder (Two-Tower) vector search. Code: https://github.com/Sunnidhya/Distilling-Lexical-Product-Associations-into-Deep-Transformers.
Read the full abstract

Computer Vision

Cross-Modal Contrastive Learning from Histopathology and CT for Automated Renal Cell Carcinoma Grading

Amit Das, Tanmay Shukla, Naofumi Tomita et al. (+9 more)

Background: Clear cell renal cell carcinoma (ccRCC) exhibits substantial clinical heterogeneity, and accurate grade assessment is essential for risk stratification and treatment planning. However, conventional grading requires invasive tissue sampling. We developed RCC-Align, a cross-modal contrastive learning framework that leverages paired histopathology and computed tomography (CT) data during training to improve noninvasive CT-based ccRCC grade prediction. Methods: RCC-Align aligns paired whole-slide histopathology images (WSIs) and CT scans through contrastive cross-modal objectives, transferring grade-discriminative information from microscopic tissue morphology to macroscopic radiologic representations. The framework was trained and evaluated on paired TCGA and CPTAC cohorts using patient-level five-fold cross-validation. Performance for low- versus high-grade ccRCC classification was compared against CT-only baselines (DINOv2-Base and DINOv2-Finetuned) and a WSI-based reference model (GigaPath-Finetuned). Cross-modal alignment was assessed using cosine similarity analysis. Results: RCC-Align achieved an AUC of 0.601 (95% CI, 0.524-0.673) and AUPRC of 0.599 (95% CI, 0.541-0.676), outperforming DINOv2-Finetuned (AUC 0.545; AUPRC 0.543) with significantly improved low-grade prediction (p = 0.004). RCC-Align also demonstrated stronger paired WSI-CT embedding alignment compared with baselines. The WSI-based GigaPath reference achieved an AUC of 0.719. Conclusion: Pathology-guided contrastive learning improves CT-based ccRCC grading while requiring only CT at inference. This approach may complement tissue diagnosis when biopsy is unsafe, infeasible, or limited by intratumoral heterogeneity. Validation in larger, multi-institutional cohorts with external testing is needed before clinical translation.
Read the full abstract

All authors Amit Das, Tanmay Shukla, Naofumi Tomita, Faraz Farhadi, Jessica Sin, Ari Hakimi, Chad Vanderbilt, Jie-Fu Chen, Ritesh Kotecha, Weijie Ma, Bing Ren, Saeed Hassanpour

Robotics

Bend the Clock: Predicting Ahead to Beat Latency in Event-Based Object Detection

Biswadeep Sen, Benoit R. Cottereau, Nicolas Cuperlier et al. (+1 more)

Event cameras promise low-latency perception for high-speed robotic systems, where even short delays can render detections stale by the time they inform downstream robotic decisions. Yet modern event detectors still require tens of milliseconds of computation before their predictions become available. Conventional evaluation ignores this delay by comparing predictions with annotations at the observation timestamp, even though the scene may have changed by the time those predictions are produced. We study this observation-availability mismatch in event-based multi-object detection and show that state-of-the-art event detectors degrade substantially when evaluated at prediction availability rather than observation time. To address this, we introduce ChronoFuse, a causal availability-time detector that predicts object states for when its output becomes available rather than for when its input was observed. ChronoFuse performs causal cross-time fusion over a multi-scale feature hierarchy, combining current representations with cached temporal features to expose short-term temporal cues without using future observations. The fusion pathway is lightweight, adding only 0.17 million parameters and 0.84 ms of mean end-to-end latency overhead. ChronoFuse recovers 71% of the accuracy lost to latency on 1Mpx driving data and 90.8% under rapid drone motion on FRED, nearly restoring zero-delay performance. Under the extreme motion of EV-Flying, ChronoFuse reaches 20.95 sAP, compared with 2.25 for the strongest standard event detector (9.3x gain). These results show that predicting ahead can be critical for robots operating in fast-changing scenes, including autonomous driving, agile flight, and robotic interception.
Read the full abstract

All authors Biswadeep Sen, Benoit R. Cottereau, Nicolas Cuperlier, Terence Sim

Machine Learning

On Preference Coverage Collapse from Hindsight Relabeling in Multi-Objective Reinforcement Learning

Baptiste Bonin, Caro Strickland and Audrey Durand

Hindsight relabeling which retroactively replacing a transition's goal with the outcome the agent actually achieved is an effective tool for improving sample-efficiency in Reinforcement Learning (RL). A natural extension to preference-conditioned multi-objective RL (MORL) relabels transitions with the preference direction the agent achieved rather than the one asked for. We show that this extension is frequently harmful: across four preference-conditioned off-policy algorithms spanning two critic backbones and two preference-sampling schemes on the continuous-control MO-Gymnasium suite, it degrades 19 of 36 algorithm-environment settings by as much as four standard deviations, improves only one, and leaves the rest unaffected. The harm is not a symptom of noisy relabels; denoising the target recovers almost nothing, and neither prioritized sampling nor any buffer-structural choice reproduces it. Instead, repeated relabeling collapses the critic's coverage onto whatever narrow region of the preference space the agent happened to visit. We name this failure mode Preference Coverage Collapse, and quantify it with abandoned preference mass (APM), a value-aware statistic that tracks the harm ($\rho = -0.73$) where a purely structural coverage count does not. We then introduce her\_mix, a single-parameter convex combination pulling the achieved direction back towards the requested preference. At one fixed value across every algorithm and environment, it returns 16 of the 19 harmed settings to baseline, preserves and even improves the one setting in which relabeling helps, and cuts abandoned preference mass from $69\%$ to $6\%$. Protecting coverage over the preference simplex, not filtering noisy relabels, is what makes hindsight relabeling safe for MORL.
Read the full abstract

Earth and Planetary Astrophysics

Simulated Effects of Radioactive Elements on Magnetic Evolution and Atmospheric Mass Loss of Earth-like Planets

Nathaniel Tanglin and Juliette Becker

Thousands of exoplanets have been discovered in recent decades, yet Earth remains the only known world with life. To assess the potential habitability of other planets, we must identify the importance of various physical conditions that may contribute to sustaining life. One such factor is the presence of a planetary magnetic field. A planet's magnetic field can protect its atmosphere from photoevaporation, and its long-term strength depends on interior heat flow driven by radiogenic heating from isotopes such as 40K, 232Th, 235U, and 238U. In this work, we use the VPLanet code to simulate the coupled thermal, magnetic, and atmospheric evolution of Earth-like planets with varying radiogenic heat budgets orbiting G-, K-, and M-dwarf stars. The range of isotope abundances tested is based on observed stellar elemental distributions for planet-hosting stars in the Milky Way. We find that variations in radiogenic heating slightly extend dynamo lifetimes and reduces atmosphere and surface water loss for K- and G-dwarf systems, while its effects are more negligible for M dwarfs. Our results show that, while stellar irradiation and planetary location primarily control photoevaporation rates, radiogenic heating contributes to planetary habitability in a small subset of cases and may help refine target selection for future surveys.
Read the full abstract

Mesoscale and Nanoscale Physics

Interlayer-correlated fractional quantum Hall state in a trilayer electron system

Chengyu Wang, C. T. Tai, N. Toemtrisna et al. (+4 more)

In multilayer quantum Hall systems, when the layer separation is sufficiently small, the interplay between intralayer and interlayer Coulomb interactions can lead to exotic, multi-component, many-body states. A particular example is the even-denominator fractional quantum Hall state (FQHS) at total filling factor $\nu=1/2$ in bilayer systems with negligible interlayer tunneling. This state can be understood as a generalized Laughlin state described by the Halperin-Laughlin $\Psi_{331}$ wavefunction. While such correlated states have been extensively explored in bilayers, little is known about their counterparts in systems with more than two layers. Here, we investigate a trilayer two-dimensional electron system confined to ultrahigh-quality GaAs triple quantum wells. We observe an exotic FQHS at total filling factor $\nu=5/7$, evinced by a deep longitudinal resistance minimum and a quantized Hall plateau, when the side layers have a density larger than the middle layer and $d/l_B \simeq 2.2$ ($d$ is the interlayer distance and $l_B$ the magnetic length). This state is naturally interpreted as the long-predicted trilayer, generalized Laughlin ($\Psi_{33311}$) state, characterized by $\nu=1/3$-like intralayer correlation within each layer and $\nu=1$-like interlayer correlation between neighboring layers. Our observation establishes a new member of the Halperin-Laughlin many-body states that extends interlayer coherence to three coupled layers.
Read the full abstract

All authors Chengyu Wang, C. T. Tai, N. Toemtrisna, A. Gupta, L. N. Pfeiffer, K. W. Baldwin, M. Shayegan

Optics

Temporal Interfaces in Metamaterials with Shape-Varying Inclusions

Ishtiaque Rahman, Mohamed Hesham Mostafa and Viktar Asadchy

Temporal modulation of metamaterials is commonly achieved by varying constituent material properties, while changes in meta-atom geometry offer an additional means of controlling their effective response. Here, we develop an analytical and numerical framework for temporal scattering in metamaterials composed of cylindrical inclusions undergoing abrupt geometric transformations. Changing the radius of circular cylinders modifies the filling fraction and scalar in-plane effective permittivity, enabling temporal reflection, field amplification or attenuation, and frequency conversion. An area-preserving transformation from circular to elliptical cross sections introduces anisotropy, producing angle-dependent temporal scattering and redirecting the energy flow while conserving the wavevector. We derive these effects using effective-medium theory and temporal boundary conditions and compare the predictions with full-wave time-domain simulations. The simulations reproduce the main scattering and temporal-aiming trends, with larger angular deflections than predicted by the quasistatic model at higher permittivity contrasts. This suggests that moving toward the resonant regime may offer opportunities for stronger temporal aiming. These results identify inclusion geometry as a means of controlling wave amplitude, frequency, and energy-flow direction at temporal interfaces.
Read the full abstract

High Energy Physics (Theory)

Maximal Total Quantum Dimension at Bounded Rank in Dijkgraaf-Witten Theories

Ce Shen

We investigate the maximal total quantum dimension at a bounded categorical rank cutoff $R$ in three-dimensional Dijkgraaf–Witten theories. This optimization problem physically governs the maximal state-optimized torus topological entanglement entropy at a fixed torus ground-state degeneracy, while providing a quantitative constraint on the search space required to classify these topological order phases. We analyze how topological twists (3-cocycles) and gauge-group structures independently govern this trade-off. For elementary-Abelian gauge groups at a fixed prime $p$, we prove that introducing cocycle twists elevates the maximum total quantum dimension from square-root growth $\Theta_p(R^{1/2})$ to linear growth $\Theta_p(R)$ in the rank cutoff, driven by the geometric protection of cyclic holonomies. For untwisted theories, we establish sharp envelopes for symmetric and alternating gauge groups, and derive group-theoretic bounds across almost-simple, solvable, and radical-free families. In particular, almost-simple groups and their direct products obey a quadratic-logarithmic upper envelope: $\log\mathcal{D} = O((\log R)^2)$. Finally, we compare these scales against WZW models and discuss the status of the general quadratic-logarithmic envelope conjecture.
Read the full abstract

Computation and Language

COMED: The Missing Middle Between Routing and Collaboration in Multi-LLM Inference

Norah Alballa, Wenxuan Zhang, Salma Kharrat et al. (+4 more)

No single Large Language Model (LLM) is uniformly reliable across queries, motivating multi-model inference systems that either route among models or combine their outputs. However, routing stops after selecting an initial model, while dense collaboration invokes peers on every query. We show that collaboration is non-monotonic: peers can recover failures that no model solves alone, but can also corrupt initially correct answers. We introduce COMED (Controlled Model Escalation for Multi-LLM Deliberation), a post-anchor controller for selective cross-model collaboration. COMED uses anchor self-consistency, router margin, and a lightweight peer probe to accept confident answers, verify ambiguous cases, and escalate only when collaboration is likely beneficial. We formalize this trade-off with a rescue-harm decomposition showing that selective collaboration improves when rescued errors outweigh collaboration-induced harms. Across medical, scientific, and general reasoning benchmarks, COMED improves fixed and routed anchors in all 16 open-weight settings, with gains up to +10.7 percentage points on MedQA while invoking fewer models and using fewer decoded tokens than dense collaboration. On HLE with frontier models, COMED improves GPT-5.5 from 23.1% to 28.1%, outperforming dense collaboration and achieving the best results.
Read the full abstract

All authors Norah Alballa, Wenxuan Zhang, Salma Kharrat, Fares Fourati, Zafar Ayyub Qazi, Mohamed Elhoseiny, Marco Canini

Quantum Physics

Influence of fast and slow laser phase noise on the fidelity of the Mølmer-Sørensen trapped-ion gate

Nikita Semenin, Ksenia Khabarova and Nikolay Kolachevsky

High-fidelity two-qubit entangling gates are essential for the realization of useful quantum algorithms on quantum processors. The Mølmer-Sørensen (MS) gate has become a common choice for trapped-ion quantum computing due to its resilience to ion temperature and its demonstrated record fidelities. However, the spectral impurity of the driving laser field impacts gate performance, with phase noise influencing the qubit dynamics through mechanisms operating on different timescales. In this work, we present a comprehensive theoretical analysis of laser phase noise in the MS gate, identifying two spectral ranges which influence the gate fidelity the most: "fast" noise at frequencies near the motional mode spectrum, and "slow" noise at frequencies on the order of the inverse gate time. We derive the noise Hamiltonians for two common laser beam geometries and obtain analytical expressions for the average gate error in terms of the laser noise power spectral density and gate parameters. For slowly varying noise spectra, we provide simplified error estimates. In addition, we validate our findings against previously published numerical simulations.
Read the full abstract

Artificial Intelligence

TwinCheck: Evidence-Grounded Negative-Twin Verification for Stateful Tool Agents

Jiaxuan Dai and Tianyi Huang

A single locally plausible tool call can derail an otherwise successful agent trajectory. Suspicion alone does not justify intervention, because the replacement itself can introduce the very failure verification is meant to prevent. We introduce TwinCheck, an inference-time verification policy that considers replacement only when the trace satisfies an evidence condition tied to a trace-local failure hypothesis. It constructs a trace-grounded counterfactual alternative, a negative twin, and replaces the agent's proposal only if the twin passes structural checks and the pairwise verifier prefers it in both candidate orders. For paired evaluation, exact replay holds the agent's parsed responses and actions fixed until the first accepted replacement, separating intervention effects from resampling. In the primary analysis of 159 multi-turn BFCL V4 tasks with complete exact-replay pairs, the complete policy raises task success for GPT-5.6 Sol from 45.3% to 58.5% (95% task-bootstrap CI [8.2, 18.8]), with no observed success-to-failure regressions. Together, these findings recast execution-boundary repair as a constrained comparison, making the counterfactual action itself the object of verification.
Read the full abstract

High Energy Astrophysical Phenomena

Waiting in the wings: extended emission GRBs from high latitude emission and jet expansion

Isabelle Worssam, Benjamin P. Gompertz and Hendrik J. van Eerten

The process powering extended emission (EE) Gamma Ray-Bursts (GRBs) from compact object mergers is a debated problem due to their observed prompt emission timescales of 10 to 100s, far in excess of the expected accretion timescales following a binary neutron star (BNS) merger. Here, we present an analytical thin shell structured jet model designed to alleviate this tension. We show that a changing jet structure in which early emission is produced by a high Lorentz factor narrow, structured jet and later emission comes from a wider, lower Lorentz factor tophat-like jet can successfully reproduce the EE GRB phenomenology. All emitting shells are powered by a $0.1$ $M_{\odot}$ accretion torus and launched within a time window of 2s after merger. Spatially-resolved opacity checks confirm that our emission sites remain optically thin, in part due to the spectral softness associated with EE. We present light curves from our model compared to representative EE bursts detected by Swift: GRBs 211211A, 211227A and 060614. While our light curves provide good matches to the data, our spectra do not evolve fast enough when compared to the well-measured spectral evolution observed in GRB 211211A. We discuss an update to the model incorporating inherent spectral evolution in the shells as future work. We also show a parameter exploration of the model and identify a parameter range within which standard short GRB-like light curves could be produced, showcasing the ability of the model to produce short bursts with and without EE.
Read the full abstract

Rings and Algebras

Integral Dependence over One-Sided Ideals, Part II: G-Left Ideals and the Köthe Problem

Masood Aryapoor

We study integral dependence over left ideals using G-left ideals, namely left ideals maximal with respect to the failure of integrality of a fixed element. We establish several characterizations of G-left ideals, and show that their intersection is precisely the upper nilradical. Motivated by the Köthe conjecture, we introduce the class of left KI-rings as rings in which sums of left ideals integral over a fixed left ideal remain integral over that ideal. It is shown that left artinian rings, PI-rings, algebraic algebras, and algebras of dimension smaller than the cardinality of the ground field are left KI-rings.
Read the full abstract

High Energy Physics (Phenomenology)

Supernova Gamma-Ray Echo as an Astrophysical Near Detector for Galactic Neutrino Propagation

Garv Chauhan and Yago Porto

Core-collapse supernova neutrinos offer a unique probe of physics beyond the Standard Model over Galactic baselines. Yet even a high-statistics nearby burst may be limited by source uncertainties, especially for propagation effects that average over energy and appear only as an overall flux suppression. Inspired by near–far comparisons in terrestrial oscillation experiments, we show that next-generation MeV observations of the supernova gamma-ray echo—the delayed 511 keV signal induced by the neutrino burst in the stellar envelope—can serve as an astrophysical near detector for the emitted $\bar\nu_e$ fluence. Comparing this echo-inferred source fluence with the surviving flux measured at Earth gives a direct test of neutrino propagation between the stellar surface and the detector. For pseudo-Dirac neutrinos, this turns the fully averaged active–sterile oscillation regime from an unobservable source-normalization ambiguity into a measurable near–far deficit, with sensitivity for nearby progenitors to splittings $\delta m^2\gtrsim 10^{-18}~{\rm eV}^2(0.22~{\rm kpc}/D)(E_\star/15~{\rm MeV})$. The same near–far logic also applies to other propagation-induced modifications of the observed flux, including neutrino decay and neutrino interactions with Galactic dark matter.
Read the full abstract

Multimedia

Small Cues, Big Consequences: Learning Pivotal Cues for Multimodal Meme Classification

Akshit Sharma and Prashant W. Patil

Memes often derive their harmful, hateful, or sarcastic meaning from small but decisive visual, textual, or cross-modal cues. Existing multimodal classifiers can miss such evidence when relying mainly on global image-text representations. We introduce MemeCF, a cue-focused benchmark of 9,895 memes across harm, hate, and sarcasm, with annotations identifying the modality and rationale of the pivotal evidence. We also propose MemePIVOT, a local-global architecture for meme classification. MemePIVOT uses frozen CLIP features, unbalanced optimal transport to align words with image patches while allowing irrelevant evidence to remain unmatched, and an evidential fusion head to combine local grounding with global meme context under uncertainty. Experiments on HarMeme, PrideMM, and MemeCF show consistent gains over strong text-only, image-only, multimodal, and vision-language baselines. Cross-dataset and ablation results further show that explicit pivotal-evidence modeling improves robustness and contributes meaningfully beyond global multimodal representations. Our code and dataset are publicly available at https://github.com/AkshitSharma1/MemePIVOT
Read the full abstract