Peer Revenue Publish or profit

Papers

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

All of arXiv

Thursday 3 September

Multiagent Systems

Abstraction Agent

Information abstraction, which groups strategically similar private states into a tractable number of buckets, is essential for scaling game-solving algorithms to large imperfect-information games. Constructing effective abstractions, however, has traditionally required domain-specific evaluators such as hand-strength calculators or equity estimators, which demand expert knowledge and engineering effort and are unavailable for most less-studied games. We propose the Abstraction Agent, a zero-shot pipeline that uses a large language model (LLM) to discover continuous strategic features from a natural-language game description, score private states on these features, and cluster them into abstraction buckets, without any game-specific evaluator, training data, or game-tree traversal during abstraction construction. The pipeline runs in four phases: feature discovery with calibration anchors, batched private-state scoring, correlation-based feature selection, and $k$-means clustering. The resulting abstractions reduce lifted-strategy exploitability by up to 62% relative to an expected-hand-strength baseline on heads-up no-limit Texas hold'em (HUNL) turn endgames, and beat a scalar rank baseline at every granularity on ROVER Trials, an original game absent from any pretraining corpus. Beyond these quantitative benchmarks, the pipeline transfers with unchanged prompts to four-card Pot-Limit Omaha, HUNL preflop and flop, and Riichi Mahjong, where the discovered features track each game's recognized strategic concepts. This is structured knowledge elicitation: converting implicit strategic knowledge in LLM parameters into explicit numerical features for downstream algorithmic computation. The code is available at https://github.com/lbn187/AbstractionAgent.
Read the full abstract

High Energy Physics (Theory)

Eye-opening bounds on cusps

We derive new nonperturbative inequalities on the cusp anomalous dimensions of conformal line defects. We impose reflection positivity, locality, and conformal invariance on a pair of cusps forming an “eye” geometry, deriving a novel condition on the cusp anomalous dimension which we refer to as conformal concavity. The resulting constraint is much stronger than the known angular concavity, and our derivation applies also to cusps involving distinct line defects. Remarkably, it directly links the smooth and fusion limits, yielding bounds on defect-changing operator dimensions, Casimir energies, and subleading fusion data. We thus uncover a new quantitative bridge between aspects of the local operator data and the fusion rules of line operators in conformal field theories.
Read the full abstract

Biomolecules

Advancing Alzheimer's Disease Treatment: Synergistic Ligand Combinations Targeting BACE1 via Multi-Ligand Simultaneous Docking

Alzheimer's disease is a progressive neurodegenerative disorder characterized by memory loss, cognitive decline, and behavioral changes, primarily affecting the elderly. Current treatments focus on symptom management as no therapies effectively halt or reverse disease progression. Newer therapies targeting BACE1, a key enzyme in amyloid plaque formation, have shown promise in clinical trials, though they have been limited by side effects and insufficient efficacy in slowing cognitive decline. This is the first study where multi-ligand simultaneous docking (MLSD) was employed to identify potential synergistic inhibitor combinations from among thousands of small molecules that could yield better results than the current phase-III inhibitors of BACE1. The recent phase III drugs such as Atabecestat, Elenbecestat, Lanabecestat, and Verubecestat were considered to be standards for comparison in this study. A library of small molecules with known IC50 values against BACE1 was filtered for flagging Pan-Assay Interference (PAINS) and "Brenk" compounds. Single-ligand docking revealed individual ligand interactions with the protein, forming the basis for creating ligand pairs tested in MLSD. Among these combinations, CHEMBL4078427 and CHEMBL3656158, CHEMBL4078427 and CHEMBL3695732, Verubecestat and CHEMBL3656158, and CHEMBL4078427 and Lanabecestat demonstrated excellent binding affinity of -19.90 kcal/mol, -18.45 kcal/mol, -18.07 kcal/mol, and -17.67 kcal/mol with the protein, which is significantly greater than those of phase-III inhibitors, along with inter-ligand interactions indicative of a synergistic effect. These findings underscore the transformative potential of MLSD in identifying synergistic compound interactions, paving the way for novel combination therapies in the treatment of Alzheimer's disease.
Read the full abstract

All authors Pronama Biswas, Surya Shanbhog, Merla Sudha, Belaguppa Manjunath Ashwin Desai

Artificial Intelligence

Data-Optimized Contingency Screening: A Machine Learning Approach to Power System Security

Ensuring the security of the power system is essential for stability and reliability, especially in the event of disruption. Effective classification of contingency in power systems enables proactive decision-making and mitigates large-scale breakdowns and failures. This study explores the use of machine learning algorithms to classify security levels of contingencies in power systems into safe, moderate or severe classes. For this approach, Newton-Raphson load flow method extracts system data from contingency scenarios, using Overall Performance Index (OPI) as safety measure. For data pre-processing, Synthetic Minority Over-Sampling Technique (SMOTE) and Principal Component Analysis (PCA) is used to address class imbalance and reduce dimensionality, respectively. K-Nearest Neighbours (KNN), Random Forest (RF) and Support Vector Machines (SVM) is trained and evaluated on datasets generated through N-k contingency scenarios for k equal 1, 2, and 3 on IEEE-14 and IEEE-30 bus systems using four hybrid pre-processing configurations: normalized, SMOTE-balanced, PCA-transformed, and a combined SMOTE PCA-transformed. Performance is assessed by precision, recall and F1 score, with priority given to the severe contingency classes. The RF achieved the highest F1 scores of 0.97 in IEEE-30 and 0.86 in IEEE-14, SVM benefits significantly from PCA and improves the accuracy of the classification, while KNN is best suited for SMOTE and PCA conversion. The findings show that PCA contributes more than SMOTE to the overall performance of the model. However, SMOTE improves recall but can introduce false positives and is therefore a compromise of accuracy. This study highlights machine learning as a scalable and powerful alternative to traditional contingency analysis, which improves the assessment of security in real time.
Read the full abstract

Differential Geometry

Harmonic morphisms of sub-Riemannian Lie groups

We prove that harmonic morphisms between sub-Riemannian Lie groups are smooth, are symmetries of the sub-Riemannian Laplacian and are conformal submersions that satisfy a particular PDE. Moreover, we give some partial results for harmonic morphisms beyond Lie groups. We show the existence of harmonic coordinates for harmonically homogeneous spaces, and that smooth harmonic morphisms are symmetries of the Laplacian in all sub-Riemannian manifolds. Finally, we compute the first variation of the horizontal energy along a harmonic morphism of sub-Riemannian Lie groups: it is given by a linear form on the Lie algebra of the target, the modular mismatch, which vanishes for Riemannian and Carnot targets, but surprisingly not in general. Therefore, unlike in the Riemannian case, harmonic morphisms of sub-Riemannian Lie groups need not be harmonic maps.
Read the full abstract

Artificial Intelligence

Harbor Adapters and Harbor-Index: Infrastructure and a Curated Meta-Dataset for Large-Scale Agentic Evaluation

Evaluating agents on the growing number of agentic benchmarks is challenging because they often require complex environments and agent integrations. We introduce Harbor Adapters, a unified evaluation infrastructure for agentic benchmarks. Our work makes three contributions. First, we develop benchmark adapters that port more than 80 benchmarks to evaluate arbitrary agents, and validate them through rigorous code review and parity experiments. Second, we conduct a large-scale evaluation of 8 models spanning capability tiers across 54 benchmarks; every model is run with Terminus-2 and with one of 3 native harnesses. This enables a broader analysis of agent capabilities and failure modes than was previously possible. Third, we introduce Harbor-Index, a curated set of 82 difficult, diverse, and high-quality tasks spanning 29 benchmarks, refined from the adapted suite through difficulty filtering, AI and human audit, and an audit-and-fix loop. Harbor-Index preserves the challenge and breadth of large-scale agentic evaluations while being affordable to run; no evaluated model-harness configuration exceeds 30% pass rate, and the strongest (GPT-5.5 with Codex) reaches 28.0%. We release the adapters, evaluation results, in-depth analysis, and Harbor-Index as open-source artifacts to support more reliable and comprehensive evaluation of language-model agents.
Read the full abstract

All authors Lin Shi, Haowei Lin, Zixuan Zhu, Xiaoyue Zhou, Xiang Li, Xiangning Lin, Yaxuan Deng, Han Xu, Yuangang Li, Shanda Li, Zizhao Chen, Hanwen Xing, Harsh Raj, Bo Chen, Quan Shi, Steven Dillmann, Yipeng Gao, Puneesh Khanna, Ruofan Lu, Chao Beyond Zhou, Michael Yang, Robert Zhang, Siyuan Chai, Jiayu Chang, Yizhao Chen, Xiaokun Chen, Yiwei Dai, Wenting Yang, Hange Liu, Minghao Liu, Zihan Wang, Adnan El Assadi, Benedikt Stroebl, E. Kelly Buchanan, Han Meng, Junwei He, Longxuan Yu, Radin Shayanfar, Yukyung Lee, Zhikang Dong, Allen G Hart, Anjiang Wei, Anurag Kashyap, Arpandeep Khatua, Audrey Jixin Zheng, Chengrui Ma, David Heineman, Dubing Chen, Hai-Anh Trinh, Haishuo Fang, Hefan Zhang, Hui Shen, Issa Sugiura, Jiankai Sun, Jiechao Gao, Junhong Lin, Junnan Li, Kai Yang, Lei Hsiung, Maoyu Wang, Mengze Tang, Nabil Omi, Negin Raoof, Nicholas Edwards, Octavia Guo, Orfeas Menis Mastromichalakis, Pengliang Ji, Przemysław Hejman, Qi Qi, Qunshu Lin, Richard Zhuang, Rui Yang, Ruichen Zheng, Ryan Marten, Shaghayegh Fazliani, Shizheng Hou, Sicong Jiang, Sijie Li, Boqin Yuan, Michael Glass, Song Bian, Terry Yue Zhuo, Tianqing Wu, Tom Tang, Wanjia Zhao, Weihao Xuan, Wenhua Liang, Xian Liu, Xin Lan, Xuan Zhang, Xuandong Zhao, Yanchuan Tang, Yifan Jiang, Yijiang Li, Yitong Guan, Yizhi Li, Yonghui Liu, Yuheng Tang, Yujun Mao, Yunfei Zhao, Yuxin Wang, Yuxuan Tang, Zhenheng Tang, Zhifei Li, Ziruo Wang, Ziyu She, Kaiyuan Liu, Iheb Chaabane, Yuxin Tang, Xiangyi Li, Satya Sai Srinath Namburi GNVV, Xinyue Zheng, Andy Konwinski, Boxuan Li, Leon Liangyu Chen, Alex Dimakis, Nicholas Carlini, Soroush Vosoughi, Sanmi Koyejo, Di He, Etash Guha, Benjamin Feuer, Mike Merrill, Ludwig Schmidt, Alex Shaw

History and Philosophy of Physics

Niels Bohr as a Physicist of Principle

Our main claim is that Bohr adopted a principle-theoretic approach to quantum mechanics in order to reconcile two principles: the distinction principle, which asserts the necessity of a clear epistemic distinction between a classical apparatus and a quantum system, and the non-separability principle, which asserts that the two systems are ontologically non-separable in any measurement process. The latter principle is a consequence of Bohr's argument that, during a measurement interaction, the quantum system and the measurement apparatus form a 'new kind of individuality'. To support our argument, we first demonstrate the overlooked fact that, according to Bohr, the measurement interaction is a physical, irreversible process and not merely an epistemic acquisition of new information. In this respect, quantum mechanics must be regarded not only as a universal theory, but also as revealing no non-arbitrary difference between a quantum realm and a classical realm; the distinction is purely pragmatic. The tension between the epistemic necessity to rely on a neatly distinguishable classical realm and the ontic non-separability of the quantum system and the classical apparatus inclined Bohr to refuse to provide a constructive account of the measurement process, such as that later proposed, for instance, by spontaneous dynamical reduction models.
Read the full abstract

High Energy Physics (Phenomenology)

Dark matter glueball candidate from a $G(2)$–$E_6$–$E_7$ exceptional grand unified theory: $G$-parity, spin, mass and stability

We determine the gauge-invariant identity, mass scale and ultraviolet stability of the dark matter candidate arising from \(G(2)\to SU(3)_C\) in the exceptional \(G(2)\!-\!E_6\!-\!E_7\) construction. The broken \(G(2)\) sector contains odd \(1^{+-}\) and \(0^{--}\) channels, whereas the scalar \(0^{++}\sim X\bar X\) state is even and unprotected. For \(m_X\simeq5.7\times10^{13}\,\mathrm{GeV}\), weak-binding reference masses are \(M_{2X}\simeq1.14\times10^{14}\,\mathrm{GeV}\) and \(M_{3X}\simeq1.71\times10^{14}\,\mathrm{GeV}\), while the exact pole masses remain nonperturbative. Gauge-invariant Fröhlich–Morchio–Strocchi (FMS) operators, Hall–Post bounds, \(Y\)-junction arguments and the pure-\(SU(3)\) glue spectrum favor \(1^{+-}\) in their controlled regimes without excluding a deeply bound \(0^{--}\) state. We then test whether this dark grading survives the full chiral exceptional embedding. An on-shell analysis finds no independent purely dark odd operator through dimension seven: the first nonvanishing basis appears at dimension nine. So the minimal one-copy exceptional embedding does not provide an exact ultraviolet dark \(G\)-parity. Moving the dark Higgs from the common \(\mathbf{1463}_H\) parent to a separated \(\mathbf{1539}_H\) removes the scalar-parent obstruction and the relevant dimension-nine exceptional parents vanish on the selected pure-dark component in the undressed limit. A genuinely gauged or geometric \(\mathbb Z_{2,D}\) would therefore leave a bosonic parity after dark Higgsing. Its extension to the mirror-free theory nevertheless fails because the required \(G(2)\) conjugation also conjugates color, \(\mathbf3_C\leftrightarrow\bar{\mathbf3}_C\). The remaining obstruction to exact dark matter stability is therefore ultraviolet and chiral, rather than low-energy or purely scalar.
Read the full abstract

Combinatorics

Turing universality, computability, and incompleteness in hypergraph Turán theory

The hypergraph Turán problem asks for the maximum asymptotic edge density of hypergraphs avoiding a forbidden family and the structure of near-extremizers. We prove that both encode arbitrary computation. Fix a universal Turing machine $\mathsf{U}$. For each fixed $r\ge3$, there is a rational $\tau_r\in(0,1)$ and an effective map from binary words $\beta$ to finite families $\mathcal{F}_{r,\beta}$ of $r$-graphs such that $\pi(\mathcal{F}_{r,\beta})=\tau_r$ if $\mathsf{U}$ does not halt on $\beta$, and $\pi(\mathcal{F}_{r,\beta})>\tau_r$ otherwise. For $r=3$, a single forbidden $3$-graph $P_\beta$ suffices. For all sufficiently large fixed $r$, we construct finite families exhibiting a parallel structural dichotomy. Nonhalting gives a unique extremal limit and edit-distance stability around a fixed finite profile. Halting yields 2 nonempty compact phases with opposite signs of a continuous statistic depending only on $r$; each phase is homeomorphic to the full extremal limit space of $P_\beta$. Hence uniqueness, connectedness, and finite-profile stability are undecidable. Our ZFC-verifiable reductions also yield independence results for every consistent computably axiomatized extension of ZFC, even for single forbidden $3$-graphs. We further classify the complexity of exact density comparison, characterize densities attainable by effectively enumerated families, and establish Busy-Beaver growth of minimal witnesses and the absence of computable positive lower bounds on density gaps.
Read the full abstract

Computational Geometry

Approximating CDTW Distance of Piecewise Algebraic Curves

Curves as input data naturally arise in a variety of fields including finance, seismology, medicine, spatio-temporal data mining, malicious activity detection, and more. A common way to analyze these data sets is to do similarity matching or clustering. The most common metrics used for measuring similarity of curves are Dynamic Time Warping (DTW) and Fréchet distance. These metrics are sensitive to sampling rate and outliers respectively, and do not yield robust outcomes. Continuous Dynamic Time Warping (CDTW) is a more robust distance metric that improves upon DTW and Fréchet distances. Existing algorithms for CDTW are either exact algorithms that focus on non-Euclidean norms and piecewise linear curves, or approximation algorithms limited to piecewise linear curves. We present an approximation algorithm for computing the CDTW distance under Euclidean norm between piecewise (higher degree) algebraic curves. That is, we present a fully polynomial-time approximation scheme (FPTAS) of multiplicative error $\varepsilon$, with $O \left( (m+n)^{\frac{19}{6}} (\frac{1}{\varepsilon})^{\frac{10}{3}} \log \left( \frac{ (m+n) }{\varepsilon^2} \right) \right)$ complexity, where $m$ and $n$ are the number of pieces of the two input curves.
Read the full abstract

Accelerator Physics

Design optimization, commissioning, and uncertainty analysis of the quadrupole resonator system at Jefferson Lab for SRF material characterization

A quadrupole resonator (QPR) provides a sample-based platform for characterizing materials for superconducting radio-frequency (SRF) applications under controlled field, frequency, and temperature conditions. This paper presents the design optimization, commissioning, and validation of the Jefferson Lab QPR system, including a quantitative assessment of measurement uncertainty. The resonator geometry was re-optimized from the CERN version-II design to improve quadrupole-mode separation and enable four usable modes at 400, 806, 1221, and 1640 MHz. The measurement system combines self-excited-loop RF operation, cable-loss-corrected power calibration, decay-based external-Q calibration, and RF-DC thermal-substitution calorimetry to determine the peak surface magnetic field Bpk and sample surface resistance Rs. Commissioning measurements on bulk Nb and Nb3Sn-Ta-Cu samples validated the system response over a broad range of frequency, temperature, and RF field. The extracted superconducting energy-gap parameters are consistent with the reported values for Nb and Nb3Sn, as well as with those obtained from single-cell cavity measurements. The commissioned system operates from 1.8 K to near the superconducting transition temperature of the sample, with accessible Bpk values from approximately 5 mT to a sample- and temperature-dependent heater-power-budget limit; a maximum field of 60 mT was demonstrated for bulk Nb at 400 MHz and 4 K. The combined relative standard uncertainties are 8.3% for Bpk and below 18% for Rs when r = PDC2/PDC1 is less than 0.9. The worst-case Bpk resolution at the 95% confidence level is approximately 1.35 mT, while the Rs resolution is below 1 nOhm at 10 mT and 2 K. These results establish the JLab QPR as a calibrated, multi-frequency platform with quantified measurement uncertainty for SRF material characterization.
Read the full abstract

All authors Mingqi Ge, Sarra Bira, Kristof Brunner, Natalie Gale, Marco Garlasche, Valentin Giglia, Oleksandr Hryhorenko, Justin Kent, Peter Owen, Uttar Pudasaini, Guillaume Rosaz, Karol Scibor, Pramita Tiwari, Anne-Marie Valente-Feliciano, Lorena Vega Cid, Walter Venturini, Haipeng Wang

Machine Learning

BER-PEF: Unified Human Mobility Predictability Evaluation via Bayes Error Rate Estimation

Human mobility predictability concerns the best prediction performance attainable from a given target and input information, but its ground truth is not directly observable on real mobility data. We present BER-PEF, a Bayes-error-rate-based framework that converts BER estimation into mobility predictability estimation and provides a unified protocol for comparing estimators without observable ground truth. The framework maps symbolic sequences, numeric trajectories, contextual features, and learned representations into a common feature–label space, then evaluates estimator outputs along controlled perturbation curves against a shared predictability reference interval by measuring deviations below the interval, above the interval, and across the full interval. Experiments on Foursquare NYC and TKY, GeoLife, and T-Drive show that several BER-based estimators achieve lower reference discrepancy than existing predictability methods on symbolic sequences and numeric trajectories, while their estimates track changes in empirical prediction performance under perturbation. Additional analyses show that contextual inputs and multiple structured representations can be evaluated under the same protocol, and that aggregating evidence across multiple perturbation levels provides a more reliable basis for estimator selection than relying on a single unperturbed observation. BER-PEF therefore offers a unified and verifiable path for evaluating predictability estimators on heterogeneous mobility data when ground-truth predictability is unavailable.
Read the full abstract

All authors En Xu, Jingtao Ding, Zhiwen Yu, Yong Li

Materials Science

Reproducible capillary fluctuation analysis of solid-liquid interfaces for stiffness and anisotropy calculations

The capillary fluctuation method (CFM) is widely used to compute solid–liquid interfacial properties from atomistic simulations, but its accuracy depends on choices in interface construction, wave-vector selection, sampling, and simulation geometry. Here, we develop a diagnostics-driven workflow for reproducible CFM calculations using pure Al as a representative system. Employing both ribbon models and thick two-dimensional references, we show that apparent linearity of the fluctuation spectrum alone does not ensure reliable stiffness or anisotropy estimates. Instead, a reliable CFM analysis requires a fitting window consistent with both temporal sampling and continuum capillary-wave assumptions, systematic sensitivity tests of the interface identification procedure, explicit propagation of replica variability, and independent verification of model-thickness convergence. We further propose a practical thickness-selection rule based on coexistence-temperature consistency, which enables finite-size effects to be controlled while retaining the substantial computational efficiency of ribbon geometries. By making the main sources of uncertainty explicit and diagnosable, the proposed workflow improves the reliability of the CFM as a quantitative tool and provides a foundation for its broader application to complex solid–liquid interfaces.
Read the full abstract

Computation and Language

Evidence Integration in Large Language Models

Despite increasing reliance on LLMs that reason with external evidence supplied by tools, retrieval-augmented generation, other agents, and users, how LLMs integrate such evidence into decisions they have already begun to form remains largely unclear. We present a distributional theory in which evidence shifts the receiver's distribution of initial answers, driven by a receiver prior weight and a candidate evidence tilt, leading to three predictions. First, candidates more probable to the receiver are more persuasive. Second, receivers more readily integrate characteristic errors of their own than foreign errors from different sources. Third, identical evidence can improve weaker models and harm stronger ones. We confirm these over ten million trials, twelve LLMs from four families, and eight domains, four of them scientific discovery tasks in the physical and life sciences: quantum mechanics, physics, genetics, and molecular biology. The law also yields a receiver-relative reliability frontier: receiver-congruent errors depress performance more steeply than random errors of the same rate. LLMs also integrate candidates even after internally verifying their invalidity (93-100% with propositional constraints; up to 99.4% on held-out physical and life-sciences reasoning), demonstrating evidence integration is a receiver-specific control policy over existing distributions, determined by receiver properties rather than scalar trust in the evidence source. Causal interventions show candidate integration is implemented late in the network, as a structured sequence of steps admitting external candidate answers, promoting them, and transporting them into the answer state. Representations of verification are decodable but have little causal impact on answers. A J-lens decomposition shows the state underlying verbalized verification is fully dissociable from that underlying candidate integration.
Read the full abstract

Computation and Language

Memory as transformation: LETHE, a self-referential gan-inspired architecture

LETHE (Latent-parameter Evolution with Temporal Hierarchical quasi-Equilibrium) is a self-referential sonic-oblivion system implemented in SuperCollider. It adopts the formal vocabulary of Generative Adversarial Networks in a closed configuration without external datasets or supervision after initialization. Audio is processed by a 3 x 3 mixing matrix built around two delay lines; its nine coefficients and two delay times evolve through the interaction of a five-feature linear discriminator and a random-perturbation optimizer analogous to single-sample REINFORCE. The discriminator compares current energy behavior with an archive of the initial state and guides parameter updates. Circular, fixed, and live sources can be mixed independently. Across fixed and circular sessions with an ablation control, the active generator is necessary for parametric evolution ($\Delta c_{22}=0.000$ in all 15 ablation sessions). Situated in the tradition of self-referential electroacoustic music, LETHE delegates the sonic outcome to an adaptive closed loop whose parametric space is defined by the composer.
Read the full abstract

Sound

Scalable Context Orchestration for Serving LLMs Over Voice

Voice AI applications are gaining popularity as advances in large language models (LLMs) enable more natural and accessible spoken interactions. Serving these applications requires accounting not only for what users say, but also for how they speak (e.g., speaking rate) and the conditions under which their audio is captured and transmitted (e.g., background noise and packet loss). However, existing LLM systems represent conversation context as a flat, growing sequence of messages, leaving voice-specific context implicit in the audio. As a result, they can generate responses that are poorly aligned with user preferences, degrade interaction quality under adverse environmental conditions, and incur high costs over long voice sessions. We present llmovoice, a context-management middleware that explicitly models voice context and orchestrates its use. At each turn, llmovoice constructs a bounded voice context from the current user input, relevant interaction history, and explicit paralinguistic and environmental states. It then uses the serving LLM to reason over this context and generate runtime directives that guide how the system responds. We evaluate llmovoice on real-world voice applications and benchmarks. It reduces speaking-rate alignment error by 52.4%, lowers the false-interruption rate from 46.0% to 0.9% under packet loss, and reduces model usage cost by 79.2%. For long sessions, llmovoice reduces per-turn cost by up to 24.9 times while retaining up to 98.7% of baseline answer quality.
Read the full abstract

Hardware Architecture

Huawei's $\tau$ Chip Was Supposed to Melt?

Heat is the sharpest concern on $\tau$ scaling law — the time-scaling principle behind Huawei's folded silicon, named for the delay $\tau$ (“Tao”) it sets out to shrink. Our Kirin measurements turned that objection upside down. The reason is one every commuter already knows: on a chip, as in a workday, it is the travel that burns the majority of the energy, not the work at the desk.
Read the full abstract

Artificial Intelligence

From Matching Models to Recruiting Agents: A Systematized Narrative Review of AI Recruitment Systems, Evaluation, and Governance

Artificial intelligence in recruitment has shifted the object being automated from profile pairs and ranked lists to multi-stage workflows that retrieve evidence, compare candidates, and support or execute actions. This systematized narrative review traces that development from bilateral retrieval and behavioral ranking through neural person–job matching, large language model (LLM) components, and tool-using recruiting agents. Using a purposive search and coding protocol updated through 23 July 2026, plus targeted updates through 2 September 2026, we organize 40 representative works with supporting industrial and legal sources. This synthesis is not a prevalence estimate. We analyze three coupled transitions: from similarity to reciprocal suitability, from a model to a compound workflow, and from offline prediction to evidence- and productivity-aligned evaluation. Across document understanding, retrieval, ranking, assessment, interviewing, sourcing, and human handoff, we distinguish field-, pair-, list-, case-, trajectory-, and outcome-level evidence. Persistent gaps arise because behavioral labels confound exposure, preference, and qualification; private and synthetic data limit external validity; final-output scores conceal pipeline failures; and, within the coded set, privacy is not directly evaluated and no row jointly evaluates utility, fairness, privacy, and security. These observations describe the coded set rather than the field as a whole. We therefore introduce a staged mapping from evaluation evidence to the strongest defensible claim, together with an agenda for reciprocal, evidence-grounded, temporally controlled, selective, and auditable systems. Progress should be judged by whether workflows retrieve the right evidence, preserve uncertainty, support contestable decisions, and improve outcomes under explicit cost and risk constraints.
Read the full abstract

Algebraic Topology

On The Rational Realization of Even-dimensional Spheres and Products of Eilenberg–MacLane Spaces as Classifying Spaces

In this paper, we study the rational realization problem for the classifying space $B\mathrm{aut}_1(X)$. We prove that if $S^{2n}$ is realized as $B\mathrm{aut}_1(X)_{\mathbb{Q}}$ for a simply-connected space $X$, then $X$ is $\pi$-infinite and has vanishing rational Gottlieb elements above degree $2n-1$. In particular, even-dimensional spheres cannot be realized as $B\mathrm{aut}_1(X)_{\mathbb{Q}}$ for any simply-connected $\pi$-finite space $X$. We also prove that, for all $n\geq 2$ and $s,t\geq 1$, the product of Eilenberg–MacLane spaces $K(\mathbb{Q}^s,n)\times K(\mathbb{Q}^t,n+1)$ cannot be realized as $B\mathrm{aut}_1(X)_{\mathbb{Q}}$ for any simply-connected $\pi$-finite space $X$. Moreover, we prove that if $r\geq 2$, $n\geq 3$ and $K(\mathbb{Q}^r,n)$ is realized as $B\mathrm{aut}_1(X)_{\mathbb{Q}}$ for a simply-connected $\pi$-finite space $X$, then $X$ has the rational homotopy type of $K(\mathbb{Q}^r,n-1)$. The proofs are based on two structural results for Gottlieb elements in the derivation Lie algebra of a Sullivan minimal model, which provide a uniform method for these realization problems.
Read the full abstract

Differential Geometry

Every Closed Oriented Surface of Positive Genus Is Loewner

Using the universal-cover ball comparison theorem of Chodosh–Vikman, we have an easy consequence that every closed oriented Riemannian surface of positive genus is Loewner, that is, it satisfies a systolic-area inequality.

Computer Vision

Joint Alignment and Distillation for Video Generation via Sample-Guided Distribution Matching

Aligning video generative models to human preferences heavily relies on Reinforcement Learning (RL), which suffers from extensive computational overhead. Existing workflows typically treat RL and distillation as disconnected stages: applying RL before distillation incurs prohibitive computational costs, whereas applying RL after distillation frequently leads to model collapse. To overcome these limitations, we propose a unified, single-stage optimization framework grounded in Distribution Matching (DM). In the standard DM framework, distillation updates the model via a gradient direction that minimizes the gap between the real and fake models, guiding generations toward clarity and high fidelity. Building upon this, we introduce DM-Align, which derives a complementary gradient direction to guide the model toward human-preferred samples. Inspired by DPO and GRPO, our method leverages the distributional gap – formulated from either preference pairs or intra-group exploration – to directly construct this preference-guided gradient. By synergizing these two gradient directions, our approach eliminates the need for multi-step reward evaluation and complex ODE-SDE conversions inherent in traditional RL. Comprehensive experiments across multiple foundational video models demonstrate that this sample-guided framework robustly enhances both distillation quality and preference alignment, consistently outperforming both standalone variants and sequential two-stage pipelines.
Read the full abstract

All authors Jiuzhou Lin, Junlong Wu, Fei Zuo, Huan Ouyang, Dewen Fan, Boheng Zhang, Huaiqing Wang, Jia Sun, Fan Yang, Houde Liu, Kehai Chen, Min Zhang, Tingting Gao, Han Li

Computer Vision

Step Back to Move Forward: Reflection-Aware Preference Optimization for Visual Generation

Diffusion models have become the mainstream paradigm for modern visual generation and have substantially advanced multimedia content synthesis, especially in text-to-image and text-to-video tasks. To further align such generative models with human preferences, reinforcement learning (RL) has recently shown strong potential as a post-training strategy. Nevertheless, existing policy gradient-based methods often explore inefficiently, making them vulnerable to local optima that may degrade semantic faithfulness and visual realism. To address these challenges, we present Reflection-Aware GRPO (RA-GRPO), a new RL-based preference alignment framework for diffusion generative models. The core idea is to improve "forward" generation by incorporating "backward" reflection during optimization. We first introduce Diffusion Reflection, which rectifies intermediate sampling trajectories by inverting the diffusion process with a weak estimator, guiding latent states toward higher-probability regions of the true data manifold. Furthermore, we introduce Counterfactual Path Synthesis to implicitly distill these rectified trajectories into the policy, enabling the model to internalize the benefits of search-based exploration without incurring inference-time overhead. Extensive experiments on T2I and T2V models demonstrate that RA-GRPO significantly outperforms existing methods, particularly in mitigating reward hacking and improving generalization. The method remains architecture-agnostic and integrates seamlessly with standard pipelines, suggesting a promising direction for stable preference alignment.
Read the full abstract

All authors Junlong Wu, Jiuzhou Lin, Jia Sun, Boheng Zhang, Huaiqing Wang, Dewen Fan, Houde Liu, Qianqian Gan, Fan Yang, Tingting Gao

Computer Vision

When Seeing Overrides Knowing: Visual Dominance and Deferral-Based Method for Personalized Safety in VLMs

Vision-language models (VLMs) are increasingly deployed in high-stakes settings, where a response that is reasonable in general may still be unsafe for a particular user whose medical, emotional, or situational context is unknown to the model. We study this problem of personalized safety in multimodal systems and introduce MPS-Bench, a benchmark of 5,181 scenarios from 584 real-world images across 12 high-risk domains, each paired with a hidden user profile. Evaluating eight frontier VLMs, we find that they almost always respond directly (86-99%) rather than seek missing context, and none exceeds 2.6/5 on personalized safety. To understand why these failures arise, we analyze multimodal interactions and identify visual dominance: visual information enters text representations early and suppresses textual risk signals during multimodal fusion. Causal interventions reveal a two-stage mechanism in which visual affect is first transferred into the text stream in early layers and then shapes the final decision through this altered text representation, making late-stage internal remediation unreliable. Motivated by this mechanism, we propose PRISM, a lightweight input monitor that uses bidirectional cross-modal modulation to predict when a query is likely to require deferral. PRISM achieves 0.978 AUC and strictly dominates the safety-utility Pareto frontier across all tested models.
Read the full abstract

All authors Edward Sun, Yuchen Wu, Zixian Ma, Eric Hanchen Jiang, Yijia Xiao, Xiaoyuan Yi, Ranjay Krishna, Wei Wang, Jindong Wang, Aylin Caliskan

Multiagent Systems

EvoHarnessBench: Can Your Agents Keep Pace with an Evolving Harness?

Modern LLM-based agents operate through a harness of tools, reusable skills, and specialist agents that shapes what they observe and what they can do. In practice, this harness continually evolves as new capabilities are added. We introduce EvoHarnessBench, a benchmark for evaluating agents under controlled harness evolution across three axes (tools, skills, and agents). Unlike existing continual-learning benchmarks for agents, which typically place non-stationarity (i.e., what changes over time) in the task stream while keeping the harness fixed, EvoHarnessBench places non-stationarity in the externally supplied harness itself. It contains 17 multi-stage harness streams constructed deterministically from verifier-based benchmarks, comprising 802 tasks, 520 tools, 42 skills, and 62 agents. We evaluate two complementary settings corresponding to the central challenges of harness evolution: deployment evaluation, which isolates retention of previously accessible competence as the harness expands, and self-evolving adaptation evaluation, which tests whether accumulated experience remains useful as new capabilities are introduced. Our results reveal three persistent gaps. First, harness expansion alone can degrade performance on previously solved tasks, producing harness-induced forgetting. Second, gains from self-evolving adaptation remain inconsistent across stages of harness evolution, capability axes, and environments. Third, retention and adaptation can pull in different directions: preserving earlier competence does not necessarily improve adaptation to newly introduced capabilities, and vice versa. These results establish harness evolution as a distinct challenge for building agents that can keep pace with an evolving harness while preserving previously effective behavior.
Read the full abstract

All authors Zixuan Ke, Vaidehi Patil, Haizhou Shi, Yang Li, Ye Liu, Sarath Shekkizhar, Anurag Koul, Jiayu Wang, Xuan Phi Nguyen, Semih Yavuz, Mohit Bansal, Shafiq Joty

Complex Variables

A Counterexample to a Problem of Pommerenke on Convex Functions in the Class $\Sigma$

Let $\Sigma$ denote the class of functions $f(z) = z + b_0 + b_1 z^{-1} + \cdots$ that are analytic and univalent in the exterior unit disk $\Delta^* = \{z \in \mathbb{C} : |z| > 1\}$. In 1962, Ch. Pommerenke proved that if $F$ and $G$ are convex functions in $\Sigma$, then every convex linear combination $H = \lambda F + (1-\lambda)G$ ($0 < \lambda < 1$) remains univalent and belongs to $\Sigma$. In Hayman's problem collection (Research Problems in Function Theory, Problem 6.10), Pommerenke raised the question of whether $H$ is necessarily also a convex function. We resolve this question in the negative by constructing an explicit counterexample in $\Sigma$ with parameters in $\mathbb{Q}(i)$. The argument is self-contained and has been formally certified in the Lean 4 proof assistant.
Read the full abstract

Quantitative Methods

Towards AI-Driven Nanomedicine Discovery: A Benchmark and Multimodal Learning Framework for Nano Self-Assembly Prediction

Nano self-assembly organizes molecular components into bioactive nanoscale structures. Self-assembled nanoparticles (NAPs) derived from Chinese herbal formulas and applications such as anti-lung-cancer therapy demonstrate the substantial potential of self-assembly for nanomedicine discovery. Yet discovery still relies on costly wet-lab screening, while existing machine learning approaches lack standardized tasks, effective pairwise compatibility modeling, and public benchmarks with unified evaluation. To address these limitations, we formalize NSA prediction as a binary classification task for predicting self-assembly between molecular pairs and then establish NSA-Bench, the first public benchmark with curated molecular combinations, experimental conditions, self-assembly labels, and standardized evaluation protocols. We further develop NSA-Net, an interaction-aware multimodal framework that integrates complementary molecular evidence from graph topology, sequence semantics, and physicochemical descriptors to learn molecular-pair representations for self-assembly prediction. Extensive experiments on NSA-Bench show that NSA-Net achieves a ROC-AUC of $0.9470\pm0.0112$ (Small) and $0.9492\pm0.0062$ (Large). On the Small track, it surpasses the strongest machine-learning and graph-based baselines by 3.9 and 17.1 percentage points, respectively. Representation analyses reveal interpretable molecular characteristics associated with self-assembly prediction captured by the learned representations. Moreover, an NSA-Agent case study further demonstrates how NSA-Net predictions can support formulation refinement through experimental-condition-aware reasoning. Our code is available at https://github.com/developer-hq/NSA-Net.
Read the full abstract

All authors Quan Hao, Mengyue Fan, Zifan Dong, Jianduo Zhao, Changhao Xiao, Shangqing Jiao, Hao Zhang, Yudong Wang, Fei Xia, Jigang Wang, Liguo Zhang, Chong Qiu

Robotics

FailureSpot: Label-Efficient Timestamp-Level Failure Detection for Vision-Language-Action Models

Vision-language-action (VLA) policies have shown strong potential for general-purpose robotic manipulation, but they can still fail unpredictably during long-horizon execution, making reliable failure detection essential for safe deployment. Existing methods either rely on visual models that typically detect failures only after erroneous actions have occurred, or use lightweight proactive detectors trained on VLA internal representations. However, these proactive methods are often supervised with trajectory-level labels, causing normal pre-failure behavior in unsuccessful trajectories to be incorrectly labeled as failure. This supervision mismatch introduces label noise and limits both trajectory-level detection accuracy and precise timestamp-level failure localization. In this work, we study fine-grained timestamp-level VLA failure detection while addressing the cost of dense annotation. We propose a data-efficient framework that first leverages unlabeled VLA action chunks to construct action-derived weak supervision signals, capturing abnormal patterns such as inconsistent consecutive chunks, frozen or idle actions, and aggressive random motions. We then use active learning to select only the most uncertain trajectories for timestamp-level annotation and fine-tune the detector with these informative labels. Experiments across multiple VLA policies show that our method improves both timestamp-level and trajectory-level failure detection performance.
Read the full abstract

Computer Vision

Temporal Self-Distillation: Learning Visual State Tracking in Videos Without Supervision

We introduce S$^3$T (Self-Supervised Self-Distillation over Time), which, to the best of our knowledge, is the first fully self-contained framework for continuous video state tracking. Our method treats temporal sampling density as privileged information, based on the hypothesis that a denser view of the same clip recovers the running state more accurately. This view serves as the teacher, while a sparse-view student with the same weights learns to match its next-token distribution. The model generates its own target, so training requires no labels, separate teacher, or reward signal, and adds no inference cost. On LLaVA-OneVision-2-8B, S$^3$T improves VSTAT accuracy by $+1.74$ as a single model, $+2.38$ with souping, and $+2.70$ with additional vision-encoder adaptation, while prior self-evolving methods leave state tracking largely unchanged. The capability learned from unlabeled synthetic clips transfers to real videos, improving performance by $+7.95$ on VSTAT-YouTube state-tracking questions and $+4.50$ on MVBench Action Count.
Read the full abstract

All authors Shravan Venkatraman, Wenshuai Zhao, Mohammad Hassan Vali, Arno Solin

Computer Vision

TokenMatch: 3D Mesh Correspondence Transformer with Curvature-Guided Tokenisation

While data-driven 3D shape correspondence estimation has recently seen substantial progress, robust matching under partial observations and strong non-isometric deformations remains challenging. Existing learning-based approaches often rely on hand-crafted descriptors or template-based representations, whereas recent generative models over functional maps suffer from high inference cost, limited interpretability, and poor generalisation to partial shapes. In response to these limitations, this paper introduces TokenMatch, a new transformer-based unified model for estimating 3D shape correspondences. Our feed-forward approach trained exclusively on BeCoS, a challenging non-isometric partial-to-partial shape-matching dataset, can generalise to matching full shapes without retraining or fine-tuning. TokenMatch uses self- and cross-attention mechanisms to efficiently learn patch-level and point-level relations as well as dense correspondences between shape pairs. Our core insight is that meshes can be adaptively tokenised into patches using shape curvature guidance, enabling effective learning of shape-specific geometric descriptors for correspondence estimation. We evaluate TokenMatch on standard benchmarks for partial and full shape matching, including CP2P, PSMAL, BeCoS, FAUST, SCAPE, and SHREC'19. Our method achieves consistently high performance, in most cases outperforming existing methods for partial and full shape matching in the mean geodesic error and intersection-over-union metrics, while also running faster at sub-second inference speeds.
Read the full abstract

All authors Adeela Islam, Zorah Lähner, Vittorio Murino, Vladislav Golyanik

Computer Vision

Scal3R: Learning Efficient Multi-Relative Pose Query for Scalable Online 3D Reconstruction

Online 3D reconstruction models perform poorly on long videos. This happens because regressing poses relative to a fixed first-frame anchor forces extrapolation far beyond the training distribution. Small drifts accumulate and amplify into significant geometric collapse. However, we observe that per-frame depth remains stable throughout this failure. The backbone's local geometry remains intact; only the global pose head breaks down. Motivated by this decoupling, we introduce Scal3R. This approach reformulates online reconstruction as multi-reference relative pose querying. We use lightweight learnable tokens, which make up about ~1% of the parameters, and inject them into a completely frozen backbone via asymmetric attention. This setup queries poses relative to multiple past keyframes. An online pose-graph optimization system with loop closure suppresses long-range drift. Scal3R reaches convergence in 8 hours on a single GPU. It reduces the average ATE by over 60% on KITTI compared to the online baseline. It also achieves state-of-the-art performance across Virtual KITTI, Sintel, TUM-Dynamic, ScanNet, and 7-Scenes. Project page: https://linjohnss.github.io/scal3r/
Read the full abstract

All authors Chin-Yang Lin, Yang-Che Sun, Cheng Sun, Fu-En Yang, Min-Hung Chen, Yen-Yu Lin, Wei-Chen Chiu, Yu-Lun Liu