High Energy Physics (Theory)
Classification of order-two T-duality orbifolds at the SO(12) free fermionic point
Alon E. Faraggi, Stefan Groot Nibbelink and Benjamin Percival
Asymmetric orbifolds provide concrete examples of non-
geometric constructions dubbed T-folds. All Z2 point groups of six dimensional asymmetric order-two orbifolds are identified. The order-two T-fold configurations on the SO(12)
lattice are classified at the fermionic point of type II string theories for all Z2 point groups. The spectra of the models on these configurations are presented parametrically, in terms of certain generalised GSO phases. The minimal effective Hodge numbers were found to be (h_{11},h_{12})=(1,1) for six T-fold configurations. Asymmetric generalisations of the mirror
symmetry map are conjectured. The orientifoldable configurations using the basic worldsheet parity were identified within the classification. Twisted sectors corresponding to pure asymmetric twists may contain Rarita-Schwinger multiplets with
spin-3/2 states. Throughout the paper, generalised GSO projections are chosen to preserve the maximal amount of supersymmetry possible. Relaxing this, the order-two point groups were identified for which non–supersymmetric T-folds can be constructed. Since some of them may enhance to N=1 or, even, N=2 supergravities, we argue that appearance of
spin-3/2 states necessarily implies that the spectrum has reorder itself in a supersymmetric fashion and hence that the
vacuum energy vanishes.
Read the full abstract ▾
High Energy Physics (Theory)
Twisted $A_{2N}$ Argyres–Douglas theories: vertex algebras, Higgs branches, and double covers
Christopher Beem and Harshal Kulkarni
We investigate the associated vertex operator
algebras and
Higgs branches of two closely related families of generalised Argyres–Douglas theories, arising from the twisted $A_{2N}$ and twisted $D_{N+1}$ series, respectively. We propose that the $A$-series theories are realised as nilpotent Higgsings of the $D$-series, and we assemble a variety of pieces of evidence for this proposal. A consequence is that the associated vertex operator
algebras of the $A$-series are finite extensions of affine Kac–Moody vertex
algebras at (non-boundary) admissible levels, rather than those Kac–Moody
algebras themselves, and that their
Higgs branches are in many cases double covers of nilpotent orbit closures. The $\mathbb{Z}_2$ gaugings of the $A$-series theories then furnish examples of
unitary SCFTs whose associated vertex operator
algebras are precisely affine Kac–Moody
algebras at non-boundary admissible levels. We propose that these vertex
algebras are equipped with a non-standard $\mathfrak{R}$-filtration, and comment on compatibility with graded unitarity.
Read the full abstract ▾
Quantum Physics
Thermodynamic Proof of Quantumness without Structure
Florian Meier and Hayata Yamasaki
Demonstrating that a machine performs genuinely
quantum operations is a central challenge in
quantum information processing. Existing proofs of quantumness typically rely on computational tasks that are infeasible for classical machines under assumptions such as computational hardness, or explicit bounds on classical runtime,
memory, or oracle queries. Here we introduce an alternative paradigm: a thermodynamic proof of quantumness based on energy consumption. Using the search problem of Yamakawa and Zhandry with a structureless uniformly random oracle, we show that any classical machine solving the task with constant success probability must generate
entropy that grows exponentially with the security parameter. By Landauer's principle, this implies an exponential lower bound on heat dissipation and hence on the energy consumption of any cyclic classical implementation. In contrast, the corresponding
quantum prover solves the task with
polynomial query and gate complexity, giving
polynomial energy consumption under standard assumptions for
quantum implementations. Thus, in the random-oracle setting, sufficiently low energy consumption in a successful implementation certifies quantumness through a macroscopic thermodynamic observation of energy consumption.
Read the full abstract ▾
Strongly Correlated Electrons
Kitaev spin liquid in superconducting networks
Guilherme Delfino, Mehmet Dede, Dmitry Green et al. (+3 more)
We propose a realization of the Kitaev honeycomb
Hamiltonian – an archetypal
spin-
liquid model – in a superconducting metamaterial. The architecture consists of Cooper-pair boxes coupled through depleted
semiconductor–
superconductor heterostructures that do not require
spin–orbit coupling. The Cooper-pair boxes encode effective
spin degrees of freedom, while normal and anomalous virtual propagation through the
heterostructures mediate bond-directional interactions. Two key control parameters are an out-of-plane
magnetic flux and the
semiconductor Fermi energy. The former controls interference and distinguishes the bond directions, while tuning the latter close to the bottom of the band gives rise to an emergent Nambu-exchange
symmetry that enforces the required bond directionality. Through numerical calculations, we identify an operating regime with controlled corrections, with associated energy and length scales within experimental reach. These results establish a route toward
equilibrium quantum spin liquids in engineered superconducting networks.
Read the full abstract ▾
All authors Guilherme Delfino, Mehmet Dede, Dmitry Green, Michael J. Manfra, Charles M. Marcus, Claudio Chamon
High Energy Physics (Theory)
The Double Copy of the Black Ring
Damien A. Easson
We construct an exact Weyl double copy of the singly spinning five-dimensional Emparan–Reall black ring. Its complete nonlinear Weyl
curvature is the sum of two scalar-weighted Maxwell squares. One field is generated by the stationary Killing vector; the other carries topological
magnetic flux through a sphere linking the ring. Both are source-free Maxwell fields on the black-ring spacetime, with corresponding scalar partners obeying a common sourced wave equation. This extends exact five-dimensional Weyl doubling beyond the type-$\mathrm D$ and type-$\mathrm N$ examples to a black hole whose exterior contains both type-$\mathrm G$ and type-$\mathrm{I}_i$ regions. The Weyl
polynomial factorizes into two distinct irreducible quadratics, establishing genuine De Smet type $22$, thereby ruling out a reduction to a single Maxwell square.
Read the full abstract ▾
Quantum Physics
Adaptive thresholding for scalable measurement-based qubit reset
Qian Cao, Unnati Akhouri, Samuel Vizvary et al. (+3 more)
Fast, high-fidelity
qubit initialization is a key primitive for scalable
quantum information processing, but conventional reset protocols either require long relaxation times or discard information contained in continuous measurement records. We demonstrate an adaptive measurement-based reset protocol for superconducting
qubits that uses
Bayesian inference to update the
qubit-state estimate after each readout and dynamically adjust the feedback threshold. Unlike fixed-threshold or repeat-until-success (RUS) protocols, the method utilizes the full analog measurement history, enabling the reset decision to become progressively more conservative as confidence in ground-state preparation increases. Implemented with real-time FPGA-based feedback on a dispersively readout transmon
qubit, the protocol achieves a ground-state initialization fidelity of $99.44\pm0.04\%$ after a small number of reset rounds. We further compare adaptive reset with RUS strategies in multi-
qubit experiments and show that adaptive thresholding provides deterministic reset duration while maintaining a five-
qubit simultaneous-reset fidelity of $98.78 \pm 0.19\%$. These results establish
Bayesian adaptive thresholding as a practical and scalable route to fast
qubit reset in superconducting
quantum processors.
Read the full abstract ▾
All authors Qian Cao, Unnati Akhouri, Samuel Vizvary, Nissim Ofek, Wei Dai, Kater W. Murch
High Energy Physics (Theory)
A Quantum-Mechanical Model for the M5-Brane
Manuel Artime, Ralph Blumenhagen and Thomas Raml
We present a maximally supersymmetric extension of the BFSS
matrix model based on the 5-Lie
algebra ${\mathcal A}_6$. The theory is Lagrangian, promotes the BFSS 2-bracket structure constants to dynamical objects governed by a Chern-Simons-like kinetic term, and displays features expected from a theory of M5-branes. We show that it can also be obtained from the single lightcone M5-brane action on a spatial five-sphere after a truncation to the time coordinate. The resulting theory is a consistent truncation and, analogously to the $U(2)$ BFSS model, supports an interpretation of the $\mathcal{A}_6$ theory as a
quantum-mechanical model for the lowest-lying modes of the M5-brane.
Read the full abstract ▾
High Energy Physics (Theory)
Jinn-Ouk Gong, TaeHun Kim, Junghwan Lee et al. (+1 more)
We establish a direct
quantum Clausius relation for conformal fields inside the apparent horizon of a quasi-de Sitter spacetime. We compute the net heat flow across the horizon from the renormalized stress-energy tensor, and independently determine the time
evolution of the renormalized von Neumann
entropy using the replica method. At leading order in the Hubble-flow expansion, the two calculations agree exactly, without using the gravitational equations of motion. Beyond leading order, we find a scheme-independent
entropy balance that includes the corresponding Wald
entropy. Our results provide a
quantum field theoretic realization of dynamical horizon thermodynamics beyond exact de Sitter
equilibrium.
Read the full abstract ▾
All authors Jinn-Ouk Gong, TaeHun Kim, Junghwan Lee, Chang Sub Shin
Strongly Correlated Electrons
Re-entrant p-wave superconductivity in a Chern Dartboard Insulator
Yueyang Wu, Rebecca Chan, Taylor L. Hughes et al. (+1 more)
We formulate an exact theory for a 2-band
superconductor with inter-orbital pairing and band Hatsugai-Kohmoto (HK) interactions. We apply this theory to a Chern dartboard
insulator and compute the pair susceptibility and pair correlations analytically. We find that in the insulating phase previously predicted by BCS theory (small pairing strength $g$, chemical potential $\mu$ in the gap), the HK interaction can induce superconductivity, and possibly a Chern dartboard
superconductor, by lifting the doubly-occupied band and making it cross the chemical potential. From this HK-induced
superconductor, we notice that in a significant parameter range (HK interaction strength $U = 1$, $\mu = 0.5$), an
insulator-
superconductor-
insulator re-entrant transition exists. We isolate the re-entrance to a conflation between HK interactions and inter-orbital pairings.
Read the full abstract ▾
All authors Yueyang Wu, Rebecca Chan, Taylor L. Hughes, Philip W. Phillips
High Energy Physics (Phenomenology)
Bin-to-bin correlations in the extraction of proton's transverse structure
Laurent Favart, Francesco Hautmann, Aleksandra Lelek et al. (+1 more)
We investigate the determination of non-perturbative QCD contributions, associated with
proton's transverse structure, from high-precision experimental data for electroweak
boson production at the LHC. We demonstrate that these determinations are strongly affected by correlated uncertainties across transverse-momentum bins in the region of low
boson's transverse momenta. We illustrate our results through a detailed analysis of LHCb measurements of $Z$-
boson spectra at forward rapidities, and a comparison with CMS measurements at central rapidities. We further examine the influence on the extraction of the non-perturbative parameters from final-state electromagnetic radiation,
photon-induced background, large-$x$ contributions to parton distributions in the forward region, matching to perturbative next-to-leading order. We comment on the comparison of our results with those from the underlying-event tuning of parton-shower Monte Carlo event generators.
Read the full abstract ▾
All authors Laurent Favart, Francesco Hautmann, Aleksandra Lelek, Louis Moureaux
High Energy Physics (Theory)
The landscape and the swampland of O(3,3) gauged supergravities
Mariana Graña and Diego Perugini
Compactifications of string theory on group
manifolds with fluxes, performed via the Scherk-Schwarz mechanism, give rise to gauged supergravities. Using double field theory, this procedure can be made T-duality covariant, yielding a set of
geometric and non-
geometric fluxes whose string theory origin is often unclear. In this paper, we systematically analyze the complete catalog of half-maximal gauged supergravities in seven dimensions. By constructing a dictionary between orbifold twists and generalized fluxes, we identify which gaugings admit a realization as (a)symmetric orbifolds, and which do not, establishing the latter as candidates for the gauged supergravity swampland.
Read the full abstract ▾
High Energy Physics (Phenomenology)
Cosmic Birefringence and Axiogenesis
Raymond T. Co, Lawrence J. Hall, Keisuke Harigaya et al. (+1 more)
Recent analyses of Planck and ACT CMB data report a $2\text{-}4\sigma$ hint of isotropic
cosmic birefringence ($\beta \approx 0.3^\circ$), corresponding to a parity-violating rotation of polarization planes that points to potential new physics. We show that both the hint of
cosmic birefringence and the observed
baryon asymmetry can be simultaneously explained by the rotation of an
axion-like field. The minimal setup predicts a positive birefringence angle, consistent with the CMB hint, and matter and kination domination ending around a temperature of $\mathcal O(3\text{-}50)$ MeV, making the model falsifiable by astrophysical probes and potentially giving rise to characteristic gravitational-wave signals. We also introduce a mechanism that can exponentially enhance the angular momentum of the rotation generated by the Affleck-Dine mechanism.
Read the full abstract ▾
All authors Raymond T. Co, Lawrence J. Hall, Keisuke Harigaya, Taegyu Lee
Quantum Physics
Measurement-Induced Phase Transitions and Logical-Space Localization in Floquet Monitored Clifford Circuits
Hyunsoo Ha and David A. Huse
An
entanglement transition can occur without a corresponding transition in purification. We demonstrate this separation in spatially local monitored Clifford
circuits with a fixed number of layers per Floquet period, where a spatially random pattern of Clifford gates and Pauli measurements is repeated exactly in time. Rare spatial structures persist under this repetition and break the correspondence between
entanglement and purification found in conventional spacetime-random
circuits. In one spatial dimension, the late-time
entanglement produced with an initial product state remains area-law, while the system remains extensively mixed in the thermodynamic and long-time limits if the initial state is maximally mixed. In two or more dimensions, we find a transition between volume-law and area-law
entanglement while the system remains mixed on both sides of this transition. An extensive amount of
quantum information survives indefinitely in both phases and evolves unitarily despite the repeated measurements. We introduce the logical-support radius to characterize its spatial spreading and identify the volume-law and area-law phases with delocalized and localized logical dynamics, respectively. We further introduce the logical Krylov dimension, which counts the independent logical operators dynamically generated from an initially local operator. Its scaling resolves the transition and provides evidence for a new universality class of measurement-induced phase transitions.
Read the full abstract ▾
Astrophysics of Galaxies
CRIMSONS: An Online Tool for Modeling Chemical Enrichment with Stochastic IMF Sampling
Martina Rossi, Lapo Querci, Stefano Ciabattini et al. (+14 more)
Chemical enrichment is a universal process shaped by several interconnected factors, including
stellar feedback, the nature of chemical polluters, and the underlying distribution of
stellar masses. Three key ingredients are often difficult to treat simultaneously in chemical-
evolution models: the
stochastic sampling of the
stellar IMF, uncertainties in
stellar nucleosynthesis yields, and pre-enrichment from
primordial PopIII
stars. In low
star-formation rate environments, incomplete IMF sampling makes enrichment sensitive to the random presence or absence of massive
supernova progenitors. Here, we present CRIMSONS, the first online tool designed to model chemical enrichment while self-consistently incorporating these three ingredients. The framework allows users to explore different IMFs,
stellar population masses and
metallicities, and multiple
stellar yield sets. It follows the
evolution of individual
stars and tracks 30 elements(H-Zn), including contributions from SN, AGB
stars, and TypeIa SNe. The current implementation adopts a closed-box framework, neglecting gas flows and spatial mixing, thus providing a controlled setup to isolate the effects of IMF sampling,
stellar yields, and PopIII pre-enrichment. With CRIMSONS, we show that incomplete IMF sampling produces large abundance scatter, particularly in low-mass
stellar populations. Different yield prescriptions introduce element-dependent variations, while some abundance ratios remain comparatively robust across models. Finally, comparison between models including
primordial enrichment and those adopting only a
metallicity floor shows that PopIII pre-enrichment can significantly affect subsequent chemical
evolution. CRIMSONS therefore provides a controlled framework for quantifying the relative impact of
stochastic IMF sampling, nucleosynthetic uncertainties, and
primordial pre-enrichment on predicted abundance patterns
Read the full abstract ▾
All authors Martina Rossi, Lapo Querci, Stefano Ciabattini, Stefania Salvadori, Irene Vanni, Davide Massari, Edoardo Ceccarelli, Donatella Romano, Viola Gelli, Raffaele Pascale, Alice Mori, Elka Rusta, Ioanna Koutsouridou, Ása Skúladóttir, Laura Magrini, Riano Giribaldi, Jose Schiappacasse-Ulloa
Artificial Intelligence
Lean Pool: An AI-Maintained Archive of Formalized Mathematics
Vasily Ilin
Lean Pool is a repository of formalized mathematics. It is grown, maintained and optimized by AI
agents.
Geometric Topology
Vyacheslav Krushkal
We introduce collision rank and show that groups of finite collision rank are good in the sense of topological $4$-
manifold theory. Finitely generated groups of collision rank two are shown to exist outside the class of subexponentially amenable groups. Such examples are given by commutator subgroups of topological full groups of Sturmian systems, and include continuum many pairwise nonisomorphic infinite simple amenable groups.
Read the full abstract ▾
Audio and Speech Processing
SPADE: A Multilingual Dataset for Speech Partial Deepfake Detection and Localization
Yuan Tseng, Aishwarya Fursule, Andrew Zijun Ma et al. (+4 more)
Recent improvements in voice-cloning
speech generation systems raise concerns about misuse by malicious actors to impersonate others and spread misinformation. Detecting such tampering is difficult, since deepfakes in the wild may be created by different generative models in a wide range of
languages. Furthermore, the
speech audio may also only be partially modified, presenting a different and potentially more challenging task than detecting fully-synthetic
speech waveforms. To enable further research in this direction, we propose a
multilingual dataset for detection and localization of partially edited
speech samples. Our
dataset includes
speech in 12
languages, generated by up to five systems per
language, and includes both a training set as well as an evaluation
benchmark. To showcase the utility of our proposed
dataset, we train localization models of existing architectures and study generalization across three axes: across different
languages, across different
speech synthesis and editing systems, and across different acoustic environments. Our results show that localization models almost always generalize poorly to
speech edited by systems not seen during training. On the other hand, generalization to edited
speech in unseen
languages still degrades performance but to a lesser extent. We also augment our testing sets with noise to evaluate generalization across acoustic environments, and find that performance of localization models degrade significantly when tested on different acoustic conditions. All together, our results imply that existing deepfake
speech detection methods are insufficient for reliably detecting edit-based
speech deepfakes in various scenarios unseen during training. SPADE is publicly available on HuggingFace.
Read the full abstract ▾
All authors Yuan Tseng, Aishwarya Fursule, Andrew Zijun Ma, Vamshi Nallaguntla, Anderson Avila, Shruti Kshirsagar, David Harwath
Combinatorics
Maximizing the number of cliques in $K_{r+1}$-free graphs with forbidden properties
Aleyah Dawkins and Rachel Kirsch
Ferrero and Lesniak in 2018 found the maximum numbers of edges in $r$-partite non-
Hamiltonian graphs. Recently we found the maximum numbers of edges and $t$-cliques in $K_{r+1}$-free graphs (1) that are not
Hamiltonian or (2) that satisfy a condition on low-degree vertices related to Pósa's
theorem. Applying
theorem (2), here we extend
theorem (1) from Hamiltonicity to other properties. We determine the maximum numbers of edges and $t$-cliques in $K_{r+1}$-free graphs that avoid one of the following properties: traceability,
Hamiltonian-connectedness, $k$-path Hamiltonicity, $k$-Hamiltonicity, $k$-
Hamiltonian-connectedness, and $k$-connectedness. We find all extremal graphs having the maximum numbers of edges. On the way, we prove upper bounds on the numbers of edges and $t$-cliques in $K_{r+1}$-free graphs that avoid an arbitrary stable property that holds for sufficiently large complete graphs.
Read the full abstract ▾
Audio and Speech Processing
Qwen-Audio-Agent Technical Report
Chong Deng, Yunjie Ji, Yuxiang Kong et al. (+5 more)
We present Qwen-Audio-
Agent, a harness that combines full-duplex voice interaction with asynchronous task execution through a foreground-background architecture. A Frontend
Agent manages dialogue and selects between direct tool use and delegation, while a Backend
Agent carries out delegated tasks in a separate context. An Orchestration Runtime maintains task state, coordinates requests for user input and authorization, and schedules the return of results to the conversation. The runtime separates
speech interruption from task cancellation and execution completion from result delivery, allowing conversation to continue while delegated work proceeds. Environmental events and persistent
memory provide context within and across sessions. Independent adapters support integration with different frontend models, backend
agents, and clients. We instantiate the architecture in desktop assistance, intelligent cockpits, and voice customer service. On an in-house cockpit
benchmark of 134 cases, mixed execution achieves a task success rate of 91.04%, compared with 72.39% and 80.60% for the direct and all delegated configurations, respectively. In a separate latency evaluation on matched successful turns, mixed execution reduces mean task execution latency by 26.73% and 30.91% relative to these baselines, respectively. These results support the complementary use of direct tool calls for immediate operations and backend delegation for multi-step tasks.
Read the full abstract ▾
All authors Chong Deng, Yunjie Ji, Yuxiang Kong, Xiangang Li, Xu Li, Binbin Zhang, Haina Zhu, Jianheng Zhuo
Multiagent Systems
Indirect tipping: a social attack surface in AI agent populations
Ariel Flint, Luca Maria Aiello, Sara M. Constantino et al. (+2 more)
As generative AI
agents are deployed at scale, safety will depend not only on technical safeguards and individual model design, but also on collective
equilibria that determine how
agent populations process information, prioritize actions, and respond to uncertainty. Yet the same
equilibria that enable
agents to coordinate also create a social attack surface. The standard framework to assess this vulnerability is critical mass dynamics: the minimum fraction of
adversarial agents required to overturn an
equilibrium through direct competition. Here, we show that this approach risks underestimating system vulnerability by reducing the problem to the identification of singular tipping points, and ignoring indirect but potentially more efficient routes through which collective behavior can be redirected. Through experiments with populations of LLM
agents and an analytic framework that captures their collective dynamics at scale, we map critical-mass thresholds that define a directed, weighted
topology over the space of coordination
equilibria, and treat this
topology as a navigable landscape. We show that indirect tipping through intermediate stepping-stone
equilibria can reduce the committed minority required to reach an alternative state, bypass majority requirements, and make possible transitions inaccessible through direct challenges. The diversity of available alternatives and timing of the attack further reshape this landscape, creating opportunities for control as well as risks of unintended destabilization. These results show that an
equilibrium's resistance to committed intervention is not an intrinsic property but a structural feature of its competitive relations with alternative states. Securing populations of interacting AI
agents therefore requires mapping this social landscape alongside individual
agent capabilities and the technical channels through which they interact.
Read the full abstract ▾
All authors Ariel Flint, Luca Maria Aiello, Sara M. Constantino, Romualdo Pastor-Satorras, Andrea Baronchelli
Quantum Physics
Lukas Homeier, Simon Linsel, Lode Pollet et al. (+1 more)
Quantum simulators are beginning to prepare long-sought phases of matter that remain difficult to realize cleanly in materials. Yet identifying such phases remains a major challenge when their defining properties are inherently non-local and cannot be captured by conventional local measurements. Here we establish multiple-
quantum coherences (MQCs) as phase-sensitive diagnostics for gapped $\mathbb{Z}_2$
quantum spin liquids. Focusing on the extended toric code and using large-scale
quantum Monte Carlo simulations, we build a direct correspondence between the elementary anyonic excitations and the weights of different MQC sectors. MQCs thereby reveal anyon condensation across the phase transitions through characteristic signatures that remain robust against fluctuations that obscure conventional diagnostics. Subsystem-resolved MQCs further provide a bipartite
entanglement witness that distinguishes a classical loop gas from a
quantum-coherent closed-loops gas. We develop and
benchmark a practical protocol to extract MQCs based on the return fidelity of adiabatic round trips, and show that local coherence measurements retain experimentally accessible signatures of the phase transitions. Our results establish MQCs as a practical diagnostic for the excitations, phase structure and global constraints of
quantum spin liquids.
Read the full abstract ▾
All authors Lukas Homeier, Simon Linsel, Lode Pollet, Ana Maria Rey
Computation and Language
FinFIRST: Benchmarking Search Agents for Financial Information Retrieval, Sourcing and Traceability
Wenqing Wang, Haitao Xiang, Xinyi Zhao et al. (+8 more)
Financial search is a highly demanding task for LLM
agents, requiring not only a correct final answer but also temporally valid information
retrieval, authoritative source selection, entity and period
alignment, unit and definition consistency, and verifiable evidence for all conclusions. Existing
benchmarks predominantly evaluate only the final answer, making it difficult to localize errors or assess whether an answer is well-founded. To address this gap, we introduce FinFIRST (Financial Information
Retrieval, Sourcing and Traceability), the first financial
benchmark to jointly evaluate answers and supporting evidence through atomic rubrics. FinFIRST comprises 123 expert-authored tasks spanning a graduated difficulty spectrum, constructed from aggregate patterns of real-world financial scenarios through an 18-field taxonomy, a six-axis coverage blueprint, a registry of 138 financial sources, contributions from over 50 finance experts, and a six-stage quality-control pipeline. Each task is accompanied by an evidence-grounded reference package decomposed into atomic criteria across three dimensions: raw-information acquisition, source verification, and computation and answer formation. We evaluate 15 model configurations under a unified tool setting. Claude-Opus-5 achieves the highest atomic score of 87.59%, while GPT-5.6-Sol attains the highest strict pass rate of 71.54%. Computation and answer formation consistently lag behind raw-information acquisition across systems. FinFIRST retains final-answer correctness as the primary objective while making the supporting research process measurable, verifiable, and diagnosable.
Read the full abstract ▾
All authors Wenqing Wang, Haitao Xiang, Xinyi Zhao, Mingming Yin, Ying Zhong, Zhaoxin Huan, Qiheng Zhou, Jin Zhu, Xiaolu Zhang, Shi Chang, Jun Zhou
Lattice Field Theory
Experiment $\leftrightarrow$ lattice QCD: understanding high-temperature QCD matter
Bedangadas Mohanty
Relativistic heavy-ion collisions provide a unique experimental opportunity to study strongly interacting matter at extreme temperature and density, while
lattice quantum chromodynamics (QCD) offers a first-principles approach to the
equilibrium properties of such matter in the non-perturbative regime. The interplay between experiment and
lattice QCD has therefore become central to establishing the properties and phase structure of QCD matter. Selected areas where this connection is particularly informative are discussed, including the QCD equation of state and its role in hydrodynamic descriptions of heavy-ion collisions, transport properties of the
quark-
gluon plasma, conserved-charge fluctuations and their relation to experimental cumulants, and the ongoing search for a critical point in the QCD phase diagram. Particular
attention is given to the limitations involved in confronting
equilibrium lattice calculations with the finite, dynamical and experimentally constrained systems produced in heavy-ion collisions. Recent developments increasingly allow quantitative tests of QCD thermodynamics over an extended range of temperature and
baryon chemical potential. The continuing experimental programmes at RHIC and the LHC, together with future measurements at FAIR, NICA and the Electron-Ion
Collider, provide important opportunities for an increasingly close interplay between
lattice QCD, phenomenology and experiment.
Read the full abstract ▾
Number Theory
On the Euler transform and the floor function
Robert Frontczak
Let $(a_n)_{n\geq 0}$ be a sequence of numbers. We derive an expression for the Euler transform of a general binomial sum involving $a_n$ and weighted by the floor function. To demonstrate the usefulness of our approach, several examples are discussed. We rediscover some known identities and prove several new. In particular, we derive some new identities for weighted binomial sums involving harmonic numbers and central binomial coefficients. We also present new closed-forms for weighted series involving the Riemann
zeta function.
Read the full abstract ▾
Information Retrieval
GroundedGEO: Auditing the Evidence Gap in Generative Search Rankings
Yihan Xia, Huiling Fan, Kangrong Zhong et al. (+1 more)
Generative search systems rank products and services for consequential decisions, and publishers can cheaply make candidate text look relevant. Yet evidence status is not a text property but a claim-evidence relation: text-only rankers and defenses cannot separate honest detailed content from fabricated detail, creating an identifiability gap. We audit this gap with an evidence-paired
benchmark (50 e-commerce queries, 1,950 cases) and a claim-level reranker, GroundedGEO, that penalizes query-relevant claims lacking support in a supplied packet. Matched rich variants control format and volume; packet twins add attestations at fixed text, while thinned packets withdraw them. On the frozen listwise ranker Qwen2.5-7B, unsupported-rich variants show significant normalized rank gain over clean candidates (+0.065 to +0.092 across claim profiles, Holm-corrected), while supported and neutral controls do not; the effect is model-dependent (marginal on
MiMo-v2.5, absent on GLM-5.3-Flash). On a frozen pointwise scorer, oracle evidence labels cut the unsupported-rich top-3 rate from 0.65 to 0.43 (laundering from 0.61 to 0.39) at lambda=40 with zero false suppression; packet twins restore the original rates without changing text. Against a 370-claim human gold, all tested automatic judges fail the preregistered reliability gate, although the best local judge retains 79-100% of oracle suppression with zero measured false suppression on protected arms. Separately, stripping attestation coverage increases false suppression by 0.307. These diagnostic effects identify two limits on the evidence channel: label quality and packet coverage. They do not validate an automatic defense, and interpretation of the adverse human-gold arm remains pending adjudication.
Read the full abstract ▾
All authors Yihan Xia, Huiling Fan, Kangrong Zhong, Taotao Wang
Formal Languages and Automata
Causal graph rewriting
Pablo Arrighi, Marin Costes and Luidnel Maignan
We introduce causal graph rewriting, a model of computation in which local rules are applied on directed acyclic graphs in an asynchronous manner. The non-determinism arising from asynchrony is disciplined by the oriented edges, which must be understood as both computational dependencies and locality constraints—and are themselves subject to the rewriting. We illustrate the model through two examples: a particle system, and a time-dilation example—reminiscent of general relativity. We study the well-definedness and properties of induced subgraphs and graph composition, which isolate and recombine the region affected by a rewrite. We then study locality with respect to these constructions, showing how a local rewrite preserves positions, borders, and context. Our main result concerns sequential composition: locality extends from single rule applications to arbitrary valid sequences, as any local rule is automatically $*$-local and $*$-extensive. We also formalise and prove the simulation of any one-dimensional cellular automaton.
Read the full abstract ▾
Artificial Intelligence
X-Planner: Event-Structured Task Planning for Embodied Intelligence
Howard Lu, Shalfun Li, Porter Pan et al. (+30 more)
Task planning bridges high-level instructions and executable behavior in long-horizon manipulation, yet modern
Vision-
Language-Action (VLA) systems often leave this intermediate structure implicit. Existing chain-of-thought (CoT) planners also tend to rely on coarse task-level annotations or serialize long
reasoning traces
token by
token. We present X-Planner, a planning front-end that addresses both the supervision and representation of
embodied reasoning. Our planning data combine Ego, UMI, and teleoperation under a hierarchy granularity with source-dependent annotation depth. Takeover-time annotations and human-designed failures supervise ongoing error recognition. On the model side, a shared VLM backbone exposes two event-structured plan forms: a discrete interface that emits interpretable event states and a latent interface that relays continuous CoT states across staggered
Transformer depths through Staircase Decoding. A frozen latent-to-text reconstruction objective provides a semantic anchor for the latent representation. Offline two-step planning evaluation places X-Planner second among four evaluated models on both BERTScore-F1 and a judge-based Overall score. In real-
robot experiments, respectively, outperforming the evaluated baselines. These results characterize planning-text quality and downstream execution.
Read the full abstract ▾
All authors Howard Lu, Shalfun Li, Porter Pan, Cris, Lumen, Cyril, Eric Hu, Lily Li, Maeve Zhang, Rain Sun, Robert Wang, KZ Zheng, Viggo Chen, Tim Ding, Regsis Cheng, YJ Xiao, Kian, Hai Lin, Alan Song, Elise Ma, Gody Li, Victor Yao, Yohann Tang, Ingrid Yu, Jason He, James Wang, Ryan Yu, Ping Yang, Chris Pan, Vincent Chen, Roy Gan, Hao Wang, Qian Wang
Computers and Society
From Pattern Recognizers to Personalized Companions: A Survey of Large Language Models in Mental Health
He Hu, Yucheng Zhou, Qianning Wang et al. (+8 more)
The rising global prevalence of mental health conditions, together with longstanding barriers in traditional healthcare, such as limited resources, high cost, stigma, and
privacy concerns, has created an urgent need for accessible and scalable support. Large
Language Models (LLMs) have emerged as a transformative technology with strong potential to democratize mental health support through advanced natural
language understanding and generation. However, the rapidly expanding, fragmented body of work in this area lacks a coherent evolutionary narrative, making it difficult to contextualize current progress and identify future directions. This survey addresses this gap by organizing and analyzing the literature around a central thesis: the role of LLMs in mental health is evolving through three distinct, increasingly sophisticated phases. We trace this trajectory from Phase I, in which LLMs act primarily as passive Information Tools and Pattern Recognizers for assessment; through Phase II, where they function as Empathetic Conversationalists for in-the-moment, stateless interactions; to the current frontier, Phase III, which seeks Longitudinal, Personalized Companions implemented as stateful cognitive
agents. To support this framework, we systematically review core technologies,
agent architectures (Profile,
Memory,
Reasoning, and Planning), and the critical infrastructure of
datasets and
benchmarks, highlighting how their
evolution underpins this developmental path. Viewing the field through this developmental lens, we provide a comprehensive synthesis of existing work, an insightful narrative of its trajectory, and a clear roadmap for future innovation in responsible, effective, and human-centered AI for mental healthcare. A curated collection of the resources reviewed in this survey is available at our project repository: https://github.com/Emo-gml/Awesome-Mental-Health-LLMs.
Read the full abstract ▾
All authors He Hu, Yucheng Zhou, Qianning Wang, Yingjian Zou, Chiyuan Ma, Juzheng Si, Jianzhuang Liu, Zitong Yu, Laizhong Cui, Fei Ma, Qi Tian
Quantum Algebra
On Haagerup-Izumi fusion categories
Terry Gannon, Andrew Schopieray and Harshit Yadav
For every odd $N$, we construct a Haagerup-Izumi (HI) category for $\mathbb Z/N$, using Barnes double-sine functions. Equivariantization gives near-group categories of type $((\mathbb Z/N)^2,N^2)$ for every odd $N$. We classify connected separable
algebras up to Morita equivalence in such HI categories and describe their dual categories. These duals are often HI for noncyclic groups with nontrivial pointed associators. We prove the Evans-Gannon
conjecture that the modular data of the center is always a smashed sum involving a finite metric group of order $N^2+4$ and recover this group and its quadratic form directly from the data of the HI category. We also give the HI equations and reconstruction in arbitrary characteristic and prove a Frobenius
symmetry in characteristic two. Using this, we obtain the complete list of (untwisted, subfactor type) HI categories over $\mathbb{C}$ for odd $N\le 43$. Other applications include finding three pairwise non-Morita-equivalent $\mathbb Z/15$ categories whose centers have the same modular data. We show that the center of a $\mathbb Z/5$ HI category cannot be defined (i.e., admits no split ribbon form) over a cyclotomic field, disproving a
conjecture of Davidovich-Hagge-Wang.
Read the full abstract ▾
Functional Analysis
Examples of Noncompact Quasinilpotent Hankel Operators
Yuanqi Sang
For every nonzero real number $a,$ we prove that the integral Hankel operator on $L^2(\mathbb R_{+})$ with kernel $k(t)=t^{-1+ia}$ is noncompact quasinilpotent. This answers in the affirmative a question by Peller \cite{Peller1998} about the existence of noncompact quasinilpotent Hankel operators. Moreover, we obtain the corresponding Hankel operators on the Hardy space with symbols $\left((1-z)/(1+z)\right)^{ia}.$
Read the full abstract ▾