Human-Computer Interaction
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 31 August
Social media platforms increasingly use AI-based credibility indicators to help users judge misinformation. Unlike individual human-AI decision-making, these indicators are embedded in information spread: users see both an AI prediction and earlier judgments shaped by the same AI, and their own judgments may then enter the public history. Yet how to analytically characterize this process remains under-explored. We therefore introduce a social-learning lens for this setting by extending the classical Bayesian cascade model with the AI indicator as a shared public signal. The resulting Gateway condition compares the evidence from the AI prediction with users' private impressions. Through this view, we show that AI changes what public history means. Crowd agreement may reflect accumulated independent human evidence, or repeated dependence on the same AI prediction. This creates a preservation-correction trade-off: stronger reliance on AI can preserve correct predictions, but can also lock in incorrect ones by blocking corrective private impressions. We calibrate the model using human-subject data on news veracity judgments. Although the AI outperforms human users, the average user weights it below her own impression but above several peer judgments, while individual users vary from discounting the AI to relying on it enough to cascade. Simulations show that over-reliance on a weak AI is especially harmful, and that diversifying AI signals across users can better keep the crowd informative. We conclude with implications for understanding human-AI interaction in information spread and designing misinformation interventions.
Machine Learning
Tail-Replay: Escaping the Curse of Linear Attention in Prefix Caching for Hybrid LLMs
Hybrid large language models interleave full-attention layers with linear-attention layers to reduce the cost of long-context inference. This structure complicates prefix caching: full-attention key-value caches are token-addressable, whereas linear-attention layers maintain recurrent states that cannot be rolled back to arbitrary prefix boundaries. Existing hybrid prefix caching methods address this mismatch by storing recurrent-state checkpoints. As a result, token-level matches are directly usable only at positions aligned with stored checkpoints, constraining prefix reuse to a discrete set of boundaries. We present Tail-Replay, a prefix caching mechanism that enables unconstrained token-level prefix reuse in hybrid large language models. The key insight is that linear-attention mechanisms such as Gated DeltaNet can be viewed as a structured, lossy compression of the input prefix: gated recurrent updates progressively attenuate the contributions of earlier inputs. Consequently, the recurrent state of a matched prefix can be well approximated by replaying only a short, recent suffix of that prefix. Tail-Replay exploits this property by caching the exact full-attention key-value cache while omitting recurrent-state checkpoints. On a cache hit, it reconstructs the linear-attention states by replaying a short, recent suffix of the matched prefix. As a result, the reuse boundary is determined by the shared tokens rather than by recurrent-state checkpoints. We evaluate Tail-Replay on three Gated DeltaNet-based hybrid models using the LongBench and RULER benchmarks. With only a 5–10% replay budget, it retains 92.8–99.9% of full-prefill quality on LongBench and RULER. For serving efficiency, we evaluate time-to-first-token speedups across multiple matched-prefix lengths—8K, 16K, and 32K. The speedup grows with prefix length, reaching $9.1$–$14.3\times$ over full prefill at 32K.
Read the full abstract
High Energy Astrophysical Phenomena
SN 2019cqc: A Hydrogen-rich Superluminous Supernova Showing Electron-Scattering Signatures
Hydrogen-rich superluminous supernovae without narrow emission lines (SLSNe-II) are rare transients whose powering mechanisms remain debated, particularly the role of circumstellar medium (CSM) interaction. We present photometric and spectroscopic observations of SN 2019cqc to investigate its power source and progenitor mass loss. We model the lightcurve using MOSFiT with the csm and csmni models to constrain the explosion and CSM parameters. SN 2019cqc peaked at $M_g = -20.21 \pm 0.07$ and emitted $\sim 1.7 \times 10^{50}\,\mathrm{erg}$ of energy in the form of radiation. Photometric and spectroscopic modeling indicates that CSM interaction dominates in both scenarios, with best-fit CSM and ejecta masses of $\sim 4.5~ M_\odot$ and $\sim 32~ M_\odot$, respectively. The inferred CSM properties imply an extreme eruptive mass loss at a rate of $\sim 0.2--0.3 ~M_\odot\,\mathrm{yr^{-1}}$ in the years preceding the core collapse, consistent with a variable progenitor of luminous blue variable progenitor. We observe a persistent blueshifted asymmetry in the H$\alpha$ emission line. At early times ($\lesssim +110$~d), this is attributed to electron scattering with bulk Velocity of ejecta, while the profile at late epochs ($\gtrsim +402$~d onward) suggests the subsequent formation of dust in the ejecta. Additionally, we identify a distinct, short-lived feature at $\sim$ 4600 Å , likely a blend of ionized C III/N III lines powered by the interaction of the SN-shock with the extended atmosphere.
Read the full abstract
Quantum Physics
Indirect stabilization of N -level quantum systems
We study the stabilizability of a N -level quantum systems using an indirect control ar- chitecture mediated by a noisy ancilla. After outlining the general theoretical framework for arbitrary N , we focus on the two-level case (N = 2), for which we derive a complete set of necessary and sufficient conditions for asymptotic stability. Finally, we illustrate the analyti- cal results with a simple numerical example on a two-spin system, showing convergence to a Bell state for a set of random initial conditions.
Read the full abstract
Computer Vision
ScenePilot: Grow-and-Repair Policy for Text-Driven 3D Indoor Scene Generation
Text-driven 3D indoor scene generation has advanced from dataset-bound layout modeling to open-vocabulary synthesis with large language and vision-language models. Yet existing methods remain limited: one-pass generators often yield geometrically invalid layouts, heavy post-hoc optimization is costly and unstable, and prompt-only planners lack reusable layout priors for functional grouping and object relations. We propose ScenePilot, a retrieval-augmented Grow-and-Repair framework that formulates scene generation as prior-guided incremental growth with learned rectification. Given a prompt, the Hierarchical Retrieval-Augmented Planning (HRAP) module retrieves room-, group-, and anchor-level layout priors to support functional group planning. A text-driven base generator then inserts object groups sequentially, while the Reinforcement Multimodal Repair (RMR) module performs lightweight local correction after each insertion and a final global repair after completion. To train this policy, we construct SceneReverse-17k, a repair-trajectory dataset built by perturbing high-quality 3D scenes in position, rotation, and scale, then using inverse operations as executable rectification targets. The policy predicts structured move–rotate–scale actions from rendered views, scene state, retrieved priors, and edit history. By combining HRAP with RMR, ScenePilot offers an efficient alternative to one-shot generation and heavy full-scene optimization, improving physical plausibility, functional coherence, and controllability while preserving diversity.
Read the full abstract
Mesoscale and Nanoscale Physics
Singlet-doublet transitions and Josephson currents in a superconducting ring with a quantum dot
We investigate the ground state properties of a superconducting ring embedded with a quantum dot (QD) by using a variational wave-function approach. A theoretical formulation for the treatment of the finite-U Anderson impurity coupled with a superconducting ring are presented. We demonstrate singlet-doublet transitions of the ground state for this system with the QD in the mixed valence regime. It is shown that the supercurrent in the superconductor ring shows oscillations with the external enclosed magnetic flux and exhibits abrupt jumps at the singlet-doublet phase transition points.
Read the full abstract
Computer Vision
Learning to Restore More: Continual Capability Expansion for Pretrained Image Restoration Models
Image restoration models are typically trained with a fixed set of capabilities. When new restoration requirements emerge, existing solutions usually train additional models or jointly retrain the original model with both new and historical data. Instead of designing another restoration backbone, we investigate how a trained restorer can continually acquire new capabilities without forgetting those learned previously. We propose RestoreMore, a continual capability-expansion framework that preserves the pretrained restoration model as a frozen capability anchor and learns residual expansion modules for newly arriving degradations. RestoreMore introduces a capability-oriented bi-level routing mechanism at multiple feature stages. The first routing level identifies restoration capabilities relevant to the current input, while the second selects and combines a sparse set of complementary degradation experts. This design enables newly introduced tasks to selectively reuse historical restoration knowledge and progressively enriches the expert bank available for subsequent restoration tasks. Extensive experiments on a wide range of restoration benchmarks demonstrate that RestoreMore consistently acquires new restoration abilities while preserving and improving previously learned capabilities.
Read the full abstract
Classical Analysis and ODEs
The Erdélyi–Magnus–Nevai and Krasikov Conjectures for Jacobi Polynomials
Let $p_n^{(\alpha,\beta)}$ denote the Jacobi polynomial orthonormal for the weight $(1-x)^\alpha(1+x)^\beta$ on $[-1,1]$, where $\alpha,\beta\ge-1/2$, and put $S=\alpha+\beta+1$. We prove the uniform degree–parameter estimate $$ (1-x)^{\alpha+1/2}(1+x)^{\beta+1/2} \bigl|p_n^{(\alpha,\beta)}(x)\bigr|^2 \le C\max\bigl\{1,S^{1/3},S^{1/2}(n+1)^{-1/6}\bigr\}. $$ This proves, in an equivalent symmetric parametrisation, the stronger degree-sensitive conjecture proposed by Krasikov and implies the Erdélyi–Magnus–Nevai conjecture. The proof starts from Krasikov's estimate in the high-parameter quadrant and transports it to the hard edges through weighted contiguous relations whose singular endpoint terms cancel; direct hypergeometric estimates control the remaining endpoint caps. A Bessel turning-point argument shows that the intermediate factor $S^{1/3}$ in the squared estimate cannot be omitted. We also derive degree-sensitive lower bounds for Jacobi Christoffel functions and Gauss–Jacobi quadrature weights.
Read the full abstract
Computation and Language
Lazy Grounding: Attacking Search Agents with Factual Evidence
Search agents mitigate hallucination by grounding their answers in retrieved web results. However, retrieval-based approaches also introduce an attack surface: agents may cite misinformation from poisoned search corpora containing false or malicious documents. We demonstrate that, in some cases, search agents' reasoning and responses may be steered by completely factual but distracting information. We refer to this failure as lazy grounding. We expose lazy grounding by injecting nearby evidence from answer-changing rewrites of benchmark questions into the search corpora. Each document contains factual evidence that supports a neighboring rewritten question but is retrieved for the original question. Across 12 model-benchmark pairs, the attack causes the accuracy of search agents' responses to drop by 5.9 points on average and by up to 17.3 points, while inducing nearby-answer adoption in every setting. The effect is even stronger when nearby evidence appears later or is more answer-shaped. Our results show that robust search agents must defend against not only misinformation but also the misapplication of factual evidence. The code is publicly available at https://github.com/frankyzha/lazy-grounding.
Read the full abstract
Differential Geometry
Failure of the local decay formula for Arnold multiplicities under twisted Kähler–Ricci flow
Di Nezza–Guedj–Lu asked whether local Arnold multiplicities under maximal weak twisted Kähler–Ricci flow with fixed cohomology class satisfy $\lambda(\varphi_t,x)=\max\{\lambda(\varphi_0,x)-t,0\}$. We construct counterexamples with analytic initial singularities in every dimension $n\ge2$. On the Hirzebruch surface $\mathbb F_e$, $e\ge2$, the initial current $[S]+(e+1)[F]$ yields $\lambda(\varphi_t,x)=1/2-t/e$ for $x\in S\setminus F$ and $0<t<e/2$, where $S$ is the negative section and $F$ a fibre. Products with smooth stationary factors give higher-dimensional examples. The Zariski negative part of $\{S\}+\{F\}$, the reduced support class, forces extra divisor mass. For pure divisorial data with simple normal crossings support, a smooth semipositive representative of the reduced support class's positive part gives exact short-time formulas. On one surface we construct flows with identical initial current germs, background data and global initial Arnold multiplicity, but different local rates.
Read the full abstract
Robotics
Data-Centric Neuromotor Interfaces for Portable Human-Machine Interaction
Dexterous human-machine interaction requires intuitive and expressive interfaces that can be efficiently deployed on constrained edge devices. Flexible material-based neuromotor interfaces hold considerable promise, as they decode human movement intention into natural control. Although emerging flexible electronic skins enable wearable high-fidelity data acquisition, practical deployment inevitably involves trade-offs between computational resources and portability. We present a data-centric paradigm where physiological features yield fundamental separability, providing sufficient discriminative cues for recognition. A wireless, high-bandwidth system developed for collecting various electrophysiological signals, when integrated with muscle-specific electrodes, forms a surface electromyography-based interface. Exploiting highly separable data, a 2,210-parameter model achieves 94.36% accuracy across 34 gestures and can be rapidly deployed on edge devices, establishing a new thousand-parameter benchmark for dexterous decoding. The underlying data-algorithm interactions in the data-centric paradigm are further clarified, demonstrating its feasibility in real-world scenarios. This study provides a principled and validated pathway for practical deployment of reliable neuromotor interfaces.
Read the full abstract
Software Engineering
Update from Hell: Can Coding Agents Survive Hidden Breakage in Dependency Upgrades?
Modern software systems rely heavily on third-party dependencies, but upgrading those dependencies remains a costly maintenance activity. Dependency upgrades do not always preserve the function signatures, type systems, APIs, or runtime semantics assumed by existing code. Consequently, developers often need to perform source code adaptations to accommodate dependency-induced changes. However, such code-level changes are often not explicitly communicated to project maintainers, posing a significant challenge to software reliability. Meanwhile, coding agents have emerged as a new form of software development tool and are increasingly adopted by developers due to their automation capabilities. In this paper, we introduce DEPBENCH, a benchmark consisting of 203 real-world dependency-upgrade tasks across five package ecosystems spanning five language communities, each involving hidden code-level changes that require source code adaptation. We evaluate mainstream coding agents on DEPBENCH. The best completed configuration solves only 104/203 tasks (51.2%), with substantial variation across agent harnesses, models, and ecosystems, highlighting an important gap between current agent capabilities and real-world software maintenance needs.
Read the full abstract
Data Structures and Algorithms
Beating Quadratic Time–Message Trade-off in Distributed Minimum Spanning Tree Construction
We present a new distributed algorithm for computing a minimum spanning tree (MST) in the \textsf{CONGEST-KT$_{1}$} model, where messages are limited to $O(\log n)$ bits and each vertex initially knows the identifiers of its neighbors. Our algorithm exposes a two-parameter time–message trade-off: for any $0 \leq \lambda \leq \kappa \leq 1/2$, it runs in $\tilde{O}(n^{\lambda}D_G + n^{1 - \kappa - \lambda} + n^{1 - 2\kappa + \lambda} + n^{1/2})$ rounds and uses $\tilde{O}(\min\{m, n^{1 + \kappa}\})$ messages, where $n$, $m$, and $D_G$ are the number of vertices, edges, and thenetwork diameter, respectively. In particular, setting $(\kappa, \lambda) = (1/3, 1/6)$ yields an MST algorithm running in $\tilde{O}(n^{1/2} + n^{1/6}D_G)$ rounds with only $\tilde{O}(n^{4/3})$ messages. Under the mild assumption $D_G = O(n^{1/3})$, this is round-optimal while improving the best known message bound of $\tilde{O}(n^{3/2})$. More broadly, our algorithm breaks the quadratic time–message trade-off barrier $\mathrm{\# rounds} \cdot \mathrm{\# messages} = \tilde{\Omega}(n^2)$, which no previous MST algorithm in the \textsf{CONGEST-KT$_{1}$} model has been able to overcome, and it does so for almost the entire range of the diameter $D_G$. As a byproduct, we also obtain new low-message broadcast, spanning-tree, and leader-election algorithms.
Read the full abstract
Optics
Sensitivity Comparison of Microwave-Frequency and Optical Fibre Interferometry Based on State-of-the-Art Components
We compare the sensitivity of fibre interferometers to vibrations using microwave oscillators and state of the art lasers over distances up to 70 km. High spectral purity lasers outperform only above 10 Hz, highlighting microwave advantages at low frequencies and the potential of hybrid microwave optical interferometry systems
Read the full abstract
Computation and Language
AIA$^{2}$: Attribute-Agnostic Imbalance Augmentation for Subgroup Robustness
Attributes describing data content and context can induce diverse imbalance patterns that go beyond label imbalance alone. However, existing studies primarily address label imbalance while overlooking data attributes, such as topics and demographics, which can induce meaningful subgroup structure while causing model degradation on underrepresented subgroups. We propose Attribute-Agnostic Imbalance Augmentation (AIA$^{2}$), a framework for improving model robustness under varying subgroup imbalances without explicit subgroup annotations. AIA$^{2}$ automatically discovers varying imbalances via latent semantic distributions, obtains slices with both learning difficulty and subgroup imbalance deficits, and deploys a large language model (LLM) for subgroup-aware imbalance augmentation. We have evaluated AIA$^{2}$ on 5 popular corpora with rich domains and their attribute values, covering social issues and diverse topics. Results show improved performance on the lowest-performing subgroups and consistent gains over competitive baselines. Ablation studies confirm complementary contributions from each component, and additional analyses show that AIA$^{2}$ provides a practical and consistent way to improve worst-group robustness under data subgroup imbalance. Code is available at https://github.com/trust-nlp/AIA2-Subgroup-Robustness.
Read the full abstract
Materials Science
Covariant formula for the driving force for interface migration
Interfaces of crystalline materials are strongly affected by the anisotropy of interface energy. A key quantity in this context is the interface stiffness tensor, which characterizes the response of the interface energy to changes in interface orientation and plays a role in determining the driving force for interface migration. The latter is expressible via contraction of the stiffness tensor and interface curvature tensor. Practical computations involving tensors necessarily rely on their individual components. In this work, an explicit component-wise expression is derived for the contraction of the stiffness and curvature tensors, valid in arbitrary coordinate systems. Within this formulation, to get the driving force, the curvature tensor is contracted with the pullback of a stiffness-related tensor - originally defined in three-dimensional Euclidean space - onto the interface manifold. This covariant treatment establishes a physically consistent foundation for three-dimensional computational models involving interface stiffness.
Read the full abstract
Machine Learning
CateKV: On Sequential Consistency for Long-Context LLM Inference Acceleration
Large language models (LLMs) have demonstrated strong capabilities in handling long-context tasks, but processing such long contexts remains challenging due to the substantial memory requirements and inference latency. In this work, we discover that certain attention heads exhibit sequential consistency in their attention patterns, which can be persistently identified using a coefficient-of-variation-based algorithm. Inspired by this observation, we propose CateKV, a hybrid KV cache method that retains only critical token information for consistent heads, thereby reducing KV cache size and computational overhead, while preserving the majority of KV pairs in adaptive heads to ensure high accuracy. We show the unique characteristics of our algorithm and its extension with existing acceleration methods. Comprehensive evaluations on long-context benchmarks show that, while maintaining accuracy comparable to full attention, CateKV reduces memory usage by up to $2.72\times$ and accelerates decoding by $2.18\times$ in single-sample inputs, and boosts throughput by $3.96\times$ in batch scenarios.
Read the full abstract
Computer Vision
Dynamic Hub-and-Spoke Memory for Streaming Video Understanding
Streaming video understanding requires answering questions at arbitrary times over a continuously growing visual stream. The central challenge is to compactly remember long-range history while effectively retrieving question-relevant evidence. We propose Dynamic Hub-and-Spoke Memory (D-HSM), a training-free framework that represents distant history as structured textual memory while preserving the recent frames as visual tokens for fine-grained perception. Specifically, D-HSM turns selected historical video chunks into typed textual observations and stores them in an entity-centered hub-and-spoke memory, with entities as hubs and related evidence as spokes. When answering a question, D-HSM dynamically retrieves a compact question-aware memory subset, expands it through hub-and-spoke links, and combines it with the recent visual window for frozen-VLM answer prediction. Extensive experiments on both streaming and long video benchmarks show that D-HSM consistently and substantially improves VLM backbones and outperforms other state-of-the-art online and offline video understanding baselines.
Read the full abstract
Quantum Algebra
A note on the center of the queer super Yangian ${\rm Y}(\mathfrak{q}_1)$
We investigate two families of central elements of the queer super Yangian ${\rm Y}(\mathfrak{q}_1)$, arising from the constructions of Poletaeva-Serganova and from Nazarov's quantum Berezinian. We establish an explicit relation between their generating series, thereby giving an answer to the question raised by Nazarov in \cite{Na22}.
Read the full abstract
Game Theory
Mechanism Design for Facility Location Games Under a Prelocated Facility
We study the problem of locating a new homogeneous facility under a prelocated facility. Here, a set of $n$ agents is located on a real line or a circle, each of whom has her location as private information, and her cost is the (expected) distance from her location to the nearest facility. Our goal is to design mechanisms which can approximately minimize the maximum cost or the social cost while eliciting agents' private information truthfully (i.e., strategy-proof). Based on real-life scenarios, we consider the problem in two settings: the general setting where each agent can be located at both sides of the prelocated facility, and the special setting where all the agents are located at the same side of the prelocated facility. For agents on a line, in the general setting, we design the best possible deterministic strategy-proof mechanism with $2$-approximation and provide a lower bound of $1.5-\epsilon\textbf{ }(\epsilon>0)$ for any randomized strategy-proof mechanism under the maximum cost objective. For the social cost, we obtain an upper bound of $n$ for deterministic strategy-proof mechanisms and lower bounds of $1.5$ and $1.0425$ for any deterministic strategy-proof mechanism and any randomized strategy-proof mechanism, respectively. In the special setting, we further provide a randomized strategy-proof $5/3$-approximation mechanism for the maximum cost and a deterministic strategy-proof $(n-1)$-approximation mechanism for the social cost. For agents on a circle, we provide a deterministic strategy-proof 2-approximation mechanism under the maximum cost objective.
Read the full abstract
Information Retrieval
PEARL: Front-Loading Relational Chains for Multi-Hop Table Retrieval
While large language models (LLMs) have shown strong capabilities in tabular reasoning, retrieving relevant tables remains challenging due to the fragmented and relational structure of real-world data. Existing work typically relies on whole table representations that overlook cross-table semantics induced by join relationships. We propose PEARL, a training-free framework that shifts the paradigm toward vertical partitioning-based sub-table encoding. PEARL augments the retrieval corpus offline by generating multi-hop queries over pre-identified join paths and reorganizing relevant columns into vertically partitioned corpus units, enabling effective multi-table retrieval without query-time LLM inference. Experiments show that PEARL consistently outperforms existing methods, with up to +30.05% gains in R@2 on 3-hop queries. The source code is available at https://github.com/SOOB2NHO/PEARL.
Read the full abstract
Commutative Algebra
Thresholds of singularities in characteristic zero
We study properties of characteristic zero variants of Frobenius thresholds $c^J(\mathfrak a)$ inspired by the work of Epstein-McDonald-R.G. and the second author. Suppose $R$ is an excellent domain of equal characteristic zero with a dualizing complex and $\mathfrak a, J \subseteq R$ are nonzero ideals with $\mathfrak a \subseteq \sqrt{J}$. By using a log resolution of singularities $Y$ of $(R, \mathfrak a)$ as well as the derived global sections of a line bundle on $Y$, we construct two analogs of the Frobenius threshold, the Hironaka-threshold and the Koszul-Hironaka threshold. We show that these two thresholds agree for parameter ideals and are always rational numbers. We prove that these thresholds also share many of the properties of the classical Frobenius threshold, especially for parameter ideals. If $R$ is regular, we show that the set of thresholds coincides with the set of possible multiplier-ideal-jumping numbers. Finally, we show that for parameter ideals in a Kawamata log terminal ring, our thresholds coincide with the limit of the Frobenius thresholds of the mod-$p$-reductions as $p$ goes to infinity. There is even a version of this without that Kawamata log terminal hypothesis.
Read the full abstract
Robotics
CometVLA: Co-Training on an Embodied Data Pyramid towards Physical Understanding
Vision-language-action (VLA) models remain brittle in manipulation tasks that require physical commonsense. Current physical VQA data is typically disembodied and misaligned with robot action domains. Egocentric videos are used only as auxiliary pre-training. It remains unclear whether improved VLM physical understanding actually benefits downstream action generation. Therefore, we present CometVLA to close this gap. We construct CometData and CometBench, an embodied physical VQA corpus and benchmark strictly aligned with the robot's action data and embodiment. We introduce Global Action Prior (GAP) tokens, a compact learnable bottleneck that isolates task-agnostic motion regularities and lets the action head consume physical commonsense without corrupting the pre-trained VLM backbone. We co-train CometVLA across the embodied data pyramid, spanning teleoperation, simulation, egocentric trajectories, and VQA layers. On real-world manipulation tasks and RoboTwin simulation, CometVLA consistently outperforms strong VLA baselines. Correlation analysis shows that stronger VLM performance on CometBench indicates higher VLA success rates. Results demonstrate that physical understanding pre-training genuinely benefits downstream manipulation.
Read the full abstract
Cryptography and Security
Extracting Knowledge from Tools in LLM Agents
LLM agents commonly use knowledge-based tools and access their underlying files, databases, and search indexes through tool invocation. This integration improves agents' ability to provide domain-specific services but also introduces the risk of tool-mediated knowledge extraction: source content exposed to an agent for legitimate responses may be progressively recovered from its outputs, enabling reconstruction of the knowledge source behind a target tool. This paper systematically investigates this risk and identifies two challenges introduced by tool invocation: tool-selection uncertainty, where an agent may invoke a competing tool instead of the target tool, and tool-argument compression, where fine-grained query information may be lost when the agent generates tool arguments. To tackle these challenges, we propose ToolSiphon, a query-only extraction attack that introduces two complementary signals: a target-discriminative signal, implemented through Tool Contrastive Analysis, to steer queries toward the target tool; and a response-grounded factual signal, implemented through Evidence Chained Feedback, to mitigate argument compression and progressively expand extraction coverage. Across three types of knowledge-based tools and six domain-specific datasets, ToolSiphon recovers 74.3% of source records on average when coarse-grained information about non-target tools is available, with 83.2% textual recovery and 90.2% semantic similarity. Even without such information, it recovers 66.3% of source records. ToolSiphon also remains effective against representative defenses and on three real-world agent platforms.
Read the full abstract
Numerical Analysis
From multi-layered problems to multiple two-layered problems: a novel frequency-time hybrid multiple-scattering integral equation solver
This paper proposes a novel frequency-time hybrid multiple scattering (FTH-MS) integral equation solver for time-dependent wave equation problems in general multi-layered media, with remarkable scalability with respect to the number of layers $N$. In light of the finite speed of wave propagation, the new methodology provides an innovative multiple-scattering idea of re-modeling the original $N$-layered problem into a sequence of $N-1$ two-layered sub-problems, for which the main advantages lie in that (i) each sub-problem enjoys much simpler wave scattering properties compared with the complicated problem in a multi-layered medium, (ii) it enables to develop high-accuracy solver utilizing Fourier transform and frequency-domain boundary integral equation (BIE) method; and (iii) numerical evaluation of the sub-problems in each multiple scattering step can be parallelized. Both multiplicative- and additive-type strategies are developed and equivalence results, which indicate that the $M$-th order multiple scattering sums can provide equivalent representations of the solutions up to a certain time $T(M)$, are rigorously derived. Owing to the existed result of exponential convergence of the perfectly-matched-layer (PML) truncation for two-layered problem, all the sub-problems is numerically resolved by means of the FTH method based on the Fourier transform and the PML-BIE method whose numerical evaluation is addressed utilizing the Chebyshev-based rectangular-polar solver with high accuracy. Numerical examples are presented to validate the efficiency and accuracy of the proposed method.
Read the full abstract
Fluid Dynamics
Coexisting chaos and order in micro-textured elastic flows
Viscoelastic fluid flows over micro-textured surfaces - densely covered by slender protrusions like cilia lining the body airways or villi covering the intestinal epithelium - underpin essential biological processes including transport, mixing, and absorption. Despite their ubiquity, the dynamics generated by the interplay between fluid elasticity and these complex geometries remain largely unexplored. Here we combine fully resolved numerical simulations with microfluidic experiments to reveal the flow dynamics established above dense arrays of microscopic pillars (canopies) immersed in the low-Reynolds-number flow of a viscoelastic liquid. We observe that the flow above the canopy tips spontaneously develops elastic turbulence. Remarkably, the chaotic state coexists with elastic waves emerging from the coupling between fluid elasticity and the heterogeneous shear induced by the canopy geometry. These ordered fluid motions persist across a broad range of flow conditions and canopy configurations. Our results demonstrate that coherent wave propagation and elastic turbulence are complementary manifestations of viscoelastic fluid flow. Beyond their fundamental significance, these mechanisms have broad implications for transport in biological and engineered environments, and reveal how structured geometries can harness the spontaneous dynamics of viscoelastic liquids to manipulate complex flows.
Read the full abstract
Computational Geometry
Unfolding Overlaps of the Exceptional Regular Polytopes
Complex Variables
A Note on the Converse Sendov Problem
For a polynomial of degree $n$ whose zeros lie in the closed unit disk, we determine the largest possible distance from a prescribed critical point of modulus $r$ to the nearest zero. If $n$ is even, the sharp radius is $\sqrt{1-r^2}$; if $n$ is odd, the sharp radius is strictly smaller for $r\in(0,1)$ and depends on $n$. Equality cases are also determined. The proof is based on the logarithmic-derivative identity and elementary geometric considerations.
Read the full abstract
Machine Learning
BCPPO: Bachelier-Inspired Constrained Proximal Policy Optimization for Tail-Risk-Aware Safe Reinforcement Learning
Expected-cost constraints can still permit rare, high-cost events. Monte Carlo conditional value at risk (CVaR) gradients can be noisy at high confidence, whereas critics that model an outcome distribution add complexity. We propose BCPPO (Bachelier-Inspired Constrained Proximal Policy Optimization), a proximal policy optimization (PPO) method. Separately initialized cost-prediction networks (critics), trained with random sample masks, produce disagreement that marks predictions sensitive to which state-action regions occur in the training data and to critic training. A Bachelier formula for the expected amount above a reference level converts this disagreement into a smooth policy-update penalty. Gradients from this penalty do not alter the critics, so temporal-difference (TD) critic learning is unchanged. A saturation-aware controller adjusts the mean-cost penalty and stops accumulated error from growing while that penalty is clipped. Deployment retains only the policy network. The disagreement penalty is neither a tail-event probability nor a guaranteed error bound, and it provides no safety guarantee. Across 175 runs with shared tasks, costs, budgets, training steps, and evaluation seeds, no comparator attains both higher mean return and lower mean CVaR than BCPPO in any task. On Push1, BCPPO has no lower return and no higher CVaR than every comparator, with at least one strict gain. These results support a practical balance among reward, caution around cost predictions that vary across trained critics, and policy-only deployment.
Read the full abstract
Quantum Physics
Best Annealing Path of Simulated Quantum Annealing via Efficient Adiabatic Phase Transition
Nested Simulated Annealing (NSA) is proposed for accelerating the classical simulation of Quantum Annealing (QA). NSA is based on two-spin interaction and exploits the causality of spin-flip, thereby significantly reducing computational cost. And through Unitary Coupled-Cluster (UCC) method, I establish a connection between NSA and XX interaction which is one of non-stoquastic Hamiltonians and demonstrate that NSA effectively utilizes a trivial degenerate state. Furthermore, by treating spins as continuous variables, I derive an energy formulation that incorporates both the problem Hamiltonian and the perturbation Hamiltonian induced by transverse electromagnetic field and show that the annealing path characterized by the lowest local-maximum exhibits the fastest convergence. Based on the result, I present a strategy for finding optimal annealing parameters for Simulated Quantum Annealing (SQA).