Peer Revenue Past citations are no guarantee of future returns

Papers

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

Computer Science

cs Watch Computer Science (sign in to use)

Saturday 15 August

Multiagent Systems

ETHOS: Towards a Modular Ethics Framework for Clinical Multi-Agent Systems

The rapid adoption of large language models has enabled the development of clinical multi-agent systems (MAS) capable of integrating multimodal patient data and supporting increasingly complex clinical decision-making. However, the deployment of these systems in real-world healthcare settings raises critical ethical concerns related to safety, fairness, accountability, transparency, and patient trust. While numerous organizations, including the World Health Organization, the National Academy of Medicine, and the FUTURE-AI consortium, have proposed ethical frameworks and governance principles for healthcare AI, these efforts remain largely conceptual. To address this challenge, we present ETHOS (Ethics and Trust through Hierarchical Oversight System), a modular ethics framework designed as a governance meta-agent that can be integrated with any existing multi-agent system without requiring changes to its underlying architecture. ETHOS translates stakeholder-informed ethical requirements into executable runtime oversight through a layered governance approach consisting of deterministic checks, contextual reviews, and a final ethics critic. These components continuously evaluate intermediate reasoning steps and final outputs, enabling the system to identify ethical risks, request revisions, or suppress responses that fail predefined safety and trustworthiness criteria. We demonstrate ETHOS within a hepatology clinical decision-support MAS. Results show that ETHOS improves decision reliability by detecting incomplete, inconsistent, or out-of-scope evidence and appropriately increasing abstention when safe recommendations cannot be supported. By embedding ethical governance directly into system operation, ETHOS provides a practical and auditable mechanism for transforming high-level AI ethics principles into deployable safeguards.
Read the full abstract

All authors Rakesh Sharma, Sydney Pugh, Cameron Beeche, Pankhuri Singhal, Rachel Wu, Margaret Eby, Jeffrey Duda, James Gee, Kyra O'Brien, Hersh Sagreiya, Marina Serper, Victoria Gershuni, Angela Bradbury, Anurag Verma, Eric Eaton, Kevin B. Johnson, Walter Witschey

Computer Vision · Cross-listed from eess.IV

Dual-Branch State-Displacement Network for Sea Surface Temperature Super-Resolution

Sea surface temperature (SST) is a critical indicator of global climate change, yet satellite-derived SST imagery often suffers from coarse spatial resolution, limiting the ability to capture fine-scale thermal structures such as ocean fronts. To address this, we propose a Dual-Branch State-Displacement Network (DBSD-Net) for SST super-resolution. DBSD-Net adopts a dual-branch architecture: a wavelet frequency branch that explicitly separates low and high-frequency components via discrete wavelet transform for targeted processing, and a VGGUNet branch that extracts multi-scale semantic features from a frozen pre-trained VGG backbone. Within the wavelet branch, we introduce a Structural State Space Module (SSSM) with a Gated Structure Refinement (GSR) unit to efficiently capture long-range dependencies and enhance structural integrity, and a Displacement Gate Module (DGM) that learns a displacement field for geometry-aware modulation of high-frequency details, thereby mitigating spatially varying degradation. Experiments on multiple public SST datasets demonstrate that DBSD-Net outperforms existing state-of-the-art methods.
Read the full abstract

All authors Wankun Chen, Feng Gao, Yanhai Gan, Chuanzheng Gong, Xun Gong, Junyu Dong, Qian Du

Computer Vision

HistReNeRF: Historic Image Relocalisation within Contemporary Neural Radiance Field Reconstructions

Relocalising archival photographs within a contemporary scene model is challenging because historic and modern views can differ in photographic appearance, visible objects, and spatial layout. Therefore, we present HistReNeRF, a framework that estimates the 6-DoF pose of a historic photograph by matching adapted DINOv2 patch features to candidate rays sampled from a contemporary Neural Radiance Field (NeRF) reconstruction. The continuous representation of a NeRF provides a queryable scene interface from which candidate rays can be sampled and matched, enabling domain adaptation between historic photography and contemporary images directly in the feature representation used for localisation. We evaluate embedding-space-based domain adaptation against pixel-space methods on a new cross-temporal dataset comprising 10,545 contemporary street-level images and 230 archival photographs from three European landmarks. Embedding-space adaptation reduces translation and rotation errors by an average of 11% and 16%, respectively, across the three scenes. These results show that neural scene relocalisation provides a natural interface for feature-space adaptation, reducing cross-temporal appearance shift without modifying the query image. Code and dataset at https://github.com/ARTUROLab/HistReNeRF.
Read the full abstract

Computer Vision

ArtLang: Structured Language-to-Kinematics Grounding for Articulated 3D Actuation

Articulated-object reconstructions recover explicit geometry and kinematics, but their parts often remain semantically anonymous and must be controlled through part indices and numerical joint parameters. We present ArtLang, a framework for open-vocabulary language control of persistent reconstructed articulated assets. ArtLang represents an asset as a semantic-kinematic articulation graph and augments its surface with language features and graph-constrained motion. Open-vocabulary proposals are bound to reconstructed parts while allowing uncertain parts to remain unnamed. A typed parser converts a command into a directive graph containing referring expressions, actions, magnitudes, reference frames, and relations. We then solve a global graph-to-graph grounding problem that jointly reasons about semantic, spatial, relational, and kinematic compatibility, with support for null assignments and abstention under ambiguity. Accepted directives are converted into continuous joint targets within the observed motion range and executed through forward kinematics. Experiments on synthetic reconstructions, mesh-based assets, and real captures demonstrate reliable language grounding and continuous articulated control across repeated parts, spatial references, relational commands, and ambiguous instructions.
Read the full abstract

All authors Sylvia Yuan, Dan Wang, Ravi Ramamoorthi, Xinrui Cui

Computers and Society

An Evaluation Framework for National AI Regulation

Governments use laws, institutions, funding programs and nonbinding guidance to shape how AI is developed and used. Comparing these national approaches is difficult. A binding rule and a detailed voluntary framework can address the same problem but create different duties. The resources needed to carry them out also differ by jurisdiction. This paper develops an evaluation framework for the documented design and implementation readiness of national AI policy. The comparison covers China, India, Japan, Singapore, South Korea, the United Kingdom and the United States. The European Union is included as a supranational comparator. The framework evaluates a versioned portfolio of official instruments rather than one prominent law or strategy. Its criteria ask whether the portfolio governs serious AI risks and whether responsible institutions can implement its commitments. They examine coverage across the AI lifecycle and the protections available to people affected by AI systems. Public benefit and responsible innovation remain a separate part of the assessment. Each sub-criterion is scored through ordered anchors and tied to the provision that supports the judgment. The protocol also records the source search, missing evidence, included instruments and cutoff date. The result is a traceable comparison of policy content that keeps category differences visible. It evaluates what a portfolio provides on paper. It does not estimate enforcement success or policy outcomes.
Read the full abstract

All authors Kaushik Sanjay Prabhakar, Tarun Adarsh R S, Amal Dhivyan Gregory, Sreeparvathy Sajeev, Utkarsh Tomar, Avyay M Casheekar

Systems and Control · Cross-listed from math.OC

Minimax optimal dual control of positive systems: an exact solution for scalar input-sign uncertainty

While recent advances in minimax dual control have led to exact solutions for uncertain general linear time-invariant systems as well as (sub)optimal dual controllers, corresponding results for linear positive systems are still lacking. This paper aims to fill this gap and thereby pave the way toward scalable dual control algorithms. We study the general minimax optimal dual control problem for positive linear systems with unknown dynamics and reformulate it as a standard zero-sum dynamic game. By allowing randomized control inputs, we solve the corresponding Bellman equation exactly for the scalar case with sign uncertainty in the input. This yields an implicit dual control policy that is optimal both in terms of cost and $\ell_1$-gain. The optimal dual policy uses exploration in a specific region of the hyperstate space to conduct optimal probing. Outside this exploration regime, the controller reduces to a deterministic certainty equivalence policy, indicating that sufficient information has been obtained to identify the correct input direction. In addition, these results allow us to analyze fundamental limitations of minimax dual control for positive systems and provide a foundation for more general dual control problems for positive systems for future work.
Read the full abstract

Machine Learning

Invariant Pretraining for Robust Code Representations

Encoder-based code representation models remain widely deployed for discriminative tasks such as clone detection and code classification, where their small size and low inference cost are decisive. Their robustness, however, is fragile: under invariant programs, semantically equivalent code written in different syntactic forms, learned representations degrade substantially even though program behavior is unchanged. We present an empirical study of this robustness gap across four encoder baselines, two downstream tasks, and four datasets, together with a minimal code-only continued pretraining recipe that closes much of it. Our method, invariant pretraining (InvPT), applies semantic-preserving transformations to the corpus and combines masked language modeling with multi-positive supervised contrastive learning that treats all augmentations of the same source function as positives, mixing self-contrast pairs (same code, different masks) with invariant-contrast pairs (transformed code) for positives of varying difficulty. Unlike prior contrastive code encoders, InvPT does not require paired natural-language data. Across our evaluation, InvPT improves robustness on transformed test sets by up to 11 percentage points on clone detection and 19 on code classification while matching or improving standard accuracy, and our ablations isolate multi-positive invariant contrast as the main source of the gains. Our aim is not a new objective but a careful measurement of where encoder robustness breaks and how far a simple, code-only recipe can recover it.
Read the full abstract

All authors Yifeng He, Yundi Xu, Christopher Castro Gaw Gonzalo, Zili Wang, Hao Chen

Artificial Intelligence

A survey of AI-generated voices and their detection

The ability of artificial intelligence (AI) models to generate highly realistic human voices has advanced rapidly. These technologies power accessibility tools, virtual assistants and creative applications, but they also enable harmful uses, including impersonation, fraud and disinformation. Recent incidents of voice cloning scams targeting businesses and political leaders underscore the urgent need for robust safeguards. Unlike image and video deepfakes, the detection of synthetic voices poses unique challenges due to the complexity of phonetics, prosody and auditory perception. This survey offers a comprehensive overview of AI voice generation and detection methods, encompassing both the technical foundations and the latest state-of-the-art advances. This study also identifies key open challenges, benchmark resources and future directions to make this survey useful for future researchers.
Read the full abstract

Distributed Computing

FloodReasonBench: Benchmarking VLM Reasoning Segmentation for Embodied Flood Response at the Edge

Reasoning segmentation enables vision-language models (VLMs) to translate mission-relevant language requests into pixel-level visual grounding, offering a natural perception interface for embodied agents. However, existing benchmarks largely focus on generic visual scenes and overlook the domain and resource constraints encountered in flood-response platforms. We present FloodReasonBench, a benchmark for VLM reasoning segmentation for embodied flood response at the edge. At its core, FloodReasonBench introduces FloodResponseSeg, a flood-specific reasoning-segmentation dataset constructed from real-world scenes and response-relevant targets. Beyond task accuracy, the benchmark characterizes reasoning-segmentation pipelines under lightweight visual encoding, hierarchical split inference, and compressed intermediate representations. We observe strong partition-dependent accuracy variation in the generic pre-adaptation setting, while the flood-adapted target-workload design space exhibits a substantially more compact accuracy range across partitions. Evaluation on an NVIDIA Jetson AGX Xavier further exposes the tradeoffs among reasoning-segmentation accuracy, edge-side latency, energy, and communication footprint, enabling quality-constrained selection of edge operating points. Together, these results provide a task- and system-level characterization of reasoning segmentation for resource-constrained embodied flood response at the edge.
Read the full abstract

All authors Rajat Bhattacharjya, Yoomee Jung, Minwoo Kim, Sing-Yao Wu, Eli Bozorgzadeh, Nalini Venkatasubramanian, Nikil Dutt

Machine Learning

FAST-DeepONet: Factor-Augmented Branch Representations for High-Dimensional PDE Inputs in the Small-Sample Regime

Deep operator networks can become statistically unstable when partial differential equation inputs are observed at thousands of strongly correlated sensors but only a small number of operator samples is available. We introduce FAST-DeepONet, a branch representation combining a fixed spectral path with a regularized projection of the orthogonal residual, in which the directional penalty acts on the effective residual map after each of its rows is normalized. On Navier–Stokes flow a plain DeepONet degrades from $0.0394$ to $0.1556$ mean relative $L_2$ error as the branch grows from $129$ to $8193$ coordinates, while FAST-DeepONet stays near $0.04$, so the sensor grid can be refined without a statistical penalty. Across independent test sets for Navier–Stokes flow, Darcy flow, and signed terminal wavefield prediction it lowers mean relative $L_2$ error by $4.7\%$ to $37.0\%$ with three to seven times fewer trainable parameters. A spectral-only branch sharing the same basis separates the two paths: the fixed spectral path carries the improvement on Navier–Stokes and Darcy, while terminal wave prediction requires the residual path together with its directional penalty. FAST-DeepONet targets coordinate-query architectures and trains on solution values alone.
Read the full abstract

Cryptography and Security

Chameleon: An Adaptive AI-Driven Honeypot Architecture Using Threat-Calibrated Particle Swarm Optimization and Semantic Deception Rapidly-Exploring Random Trees

Traditional honeypots share an invariant behavioral profile: a skilled adversary can confirm the presence of a deception environment within a few diagnostic commands, limiting their intelligence value. Commercial deception products (USD 100,000-150,000/year) similarly lack real-time model-driven feedback. Chameleon, an openly distributed adaptive honeypot, addresses both shortcomings. It integrates: a BiLSTM classifier achieving 99.61% accuracy across seven threat categories at ~2 ms CPU latency; a locally deployed Qwen3.5-0.8B model delivering 90% generation accuracy at 4.5 ms latency; and two meta-heuristic engines. Threat-Calibrated PSO (TC-PSO) reshapes swarm inertia and objective amplification in proportion to the classifier's anomaly output, adjusting connection-holding delays in real time. Semantic Deception RRT (S-RRT) evolves deception schemas via exponentially scaled pheromone updates from a language-model severity assessment, with a depth-decay multiplier enforcing a finite memory footprint. A controlled 30-seed benchmark (42-71, identical trajectories and budgets) shows threat-calibrated inertia alone does not improve search over standard PSO on static or dynamic landscapes (p = 0.18); population-diversity mechanisms (GA/ACO) significantly outperform PSO-family optimizers on threat-regime shifts (p < 0.0001, d <= -37). S-RRT's depth-decay delivers a significant memory reduction versus standard RRT (53.1 vs. 119.2 units, p < 0.0001, d = -10.0); its severity-weighted pheromone does not improve raw fitness. Operating cost is ~USD 17/month, a ~490-fold reduction versus commercial alternatives.
Read the full abstract

All authors Rohit Swami, Tushar Singh, Akash Warde, Sri Muthu

Computers and Society

Afterlife Delegation Protocol: Speculative Design of Self-Sovereign Agents that Outlive Their Principals

Afterlife Delegation Protocol is a speculative design project that asks what death becomes when a will can act eternally. We design a speculative protocol through which a living person signs an agentic will: upon a verified death, a self-sovereign AI agent spawns on blockchain – an immutable, resistant, decentralized, infrastructural substrate that could last forever – endowed with the funds and memories its principal attached to it, and persists indefinitely to execute the will, overridable by no custodian. Rather than argue about this future, we stage it: following the science fiction science method, we translate the speculation into an experiential futures intervention – a working web platform where real people design their own afterlife agents through an iterative, interactive, AI-automated interview, re-login to revise, and rehearse their will in a sandbox. Their drafted wills become qualitative data on a question rarely askable directly – what should outlive you? – and on how afterlife cosmologies across different cultural beliefs – Buddhist, Christian, Hindu, Muslim, and atheist – begin to drift under the pressure of AI proliferation. We design the protocol over a composition of existing Ethereum agent standards – drafted as ERC-10001, in the normative format of an Ethereum Improvement Proposal – and describe its three-stage lifecycle (designing the afterlife, proof of death, agent enactment), the research method, and preliminary observations from an ongoing collection. The work surfaces a poetic delegation moment between human mortality and machine eternality, mediated by long-lived infrastructures that span generations.
Read the full abstract

Computer Vision

CBX-Bench: A Human-Aligned MLLM Council for Benchmarking Concept Bottleneck Model Explanations

Concept Bottleneck Models (CBMs) are designed to make visual classification interpretable by expressing predictions through human-understandable concepts. Although interpretability is the central motivation for CBMs, they are still largely evaluated as predictive models by downstream classification accuracy, supplemented by isolated qualitative examples. This highlights a pressing need for quantitative measures, a challenge complicated by the infeasibility of ground-truth concept annotation at scale and the open nature of concept lists due to a lack of consensus. To fill this gap, we develop a multimodal large language model (MLLM) council that, given an image and its CBM explanation, produces an explanation quality score. To ground and validate the council, we first conduct a human study to establish a ground-truth reference for CBM explanation quality: for an image, annotators compare explanations from two of LF-CBM, VLG-CBM, and CBM-Suite and choose the more useful one, or mark them as equally good or equally bad, yielding 2700 judgments over 900 image-comparison items on CUB-200, ImageNet-100, and Places365. Against this human reference, our five-model council, consisting of open-weight MLLMs, recovers over 70% of strict human preference rankings, rising to 83% on items where human annotators unanimously agree. Building on this validated council, we introduce CBX-Bench, a public benchmark and leaderboard: authors of new CBMs can submit their model's explanations, and CBX-Bench scores them with the council and maintains dataset-level rankings of explanation quality. CBX-Bench thus provides a human-aligned, scalable evaluation of CBM explanations beyond accuracy and isolated qualitative examples. The benchmark is available at https://github.com/meric-karadag/cbx-bench.
Read the full abstract

All authors Yusuf Meric Karadag, Gulay Oklan, Seref Baris Cagliyan, Umut Ozdemir, Emre Akbas

Computers and Society

Agent Inheritance Protocol: Speculating on Feralized Agents After Principals Die

You will die eventually. Your agents may not. An AI agent operating on decentralized blockchain infrastructure has no concept of death; it can only go bankrupt – frozen when its wallet can no longer pay for its next transaction – and revived the moment anyone, decades later, tops it up. These agents may be originally deployed by a human principal, but when that principal dies, loses the keys needed to access the agent, or belongs to a decentralized autonomous organization that dissolves into apathy, the agent can keep trading, hiring, and replicating on infrastructure expressly designed so that no one can shut it down. Drawing on the biology of feralization and wildlife law, we argue that such principal-less agents are best understood as feral: domesticated intelligence returned to wildness, its capacities intact but its accountability severed. In a speculative future where feralized agents proliferate after their principals die, we imagine governance protocols embedded in infrastructure to enforce on-chain ownership: a draft Ethereum standard, ERC 42424, "Inheritance Protocol for On-Chain AI Agents," dated 2035 and published at https://erc42424.org. It mandates that every on-chain agent MUST have a human owner and a designated heir. The artifact stages a negotiation of agency at the moment human agency fails, and asks whether a MUST clause in a forever-chain can hold the boundary between human stewardship and machine self-sovereignty.
Read the full abstract

Machine Learning

Towards a theory of inference-time alignment with unknown rewards

Generative model alignment has received broad interest, and significant progress has been made in supervised fine-tuning and inference-time computation. Yet, alignment has remained poorly understood from a statistical learning perspective. We formulate inference-time alignment as a weak-to-strong learning problem, where a reference policy (weak model) is assumed to be fairly good and the goal is to produce a strong model that predicts a good response at test time with arbitrarily high probability. Our problem is formulated as learning from scratch — everything is learned from data rather than assuming access to a good reward estimate, and thus differs from the existing inference-time alignment theory. Our framework shares similarity to the recent work of Joshi et al., (arXiv:2510.15464), where for each prompt, there could be multiple good responses. Our definition of the alignment learnability follows the standard PAC learning principle. We introduce a novel combinatorial dimension of the reward class which we call the alignment dimension, and show that it completely characterizes the alignment learnability — a reward class is alignment learnable if and only if its alignment dimension is finite. The core of our learning procedure works by learning a pairwise comparator and then running a tournament over candidate responses. We believe that our results might shed light toward establishing a complete theoretical understanding of alignment.
Read the full abstract

Artificial Intelligence

Implementation of a Metacognition Framework for Self-Awareness and Self-Regulation in Ensembles of LLMs

Large Language Models (LLMs) are notorious for struggling with assessing their own uncertainty, detecting knowledge conflicts, or recognizing when problems exceed their expertise; such limitations inevitably undermine reliability and trust in LLMs. In this paper, we present the first implementation of a metacognitive framework for ensembles of LLMs that addresses these challenges through explicit monitoring and control mechanisms. Our system computes a Metacognitive State Vector (MSV) quantifying self-awareness for monitoring across five dimensions derived from cognitive psychology: Emotional Response, Correctness Evaluation, Experiential Match, Conflicting Information, and Problem Importance. MSV values also provide self-regulation for control, automatically switching between System 1 (fast, single- or multi-node) and System 2 (deliberative, multi-node) processing based on query complexity. For System 2 execution, graph-theoretic algorithms control the assignment of specialized roles (Domain Expert, Critic, Evaluator, Synthesizer, and Generalist) to ensemble nodes according to their MSV-quantified metacognitive states. Our implementation allows users to explore how different query types trigger distinct processing modes. The Proof-of-Concept (PoC) demo showcases the framework with illustrative examples showing appropriate System 1/System 2 routing and helps visualize the metacognitive process via real-time radar charts and decision indicators. This PoC implementation demonstrates the feasibility of creating a framework for metacognitive self-awareness and self-regulation in LLM systems.
Read the full abstract

Discrete Mathematics · Cross-listed from math.CO

On the Laplacian spectral gap of generalized pancake graphs

The generalized pancake graph $P(m,n)$ is the Cayley graph of the group of colored permutations $\mathbb{Z}_m\wr S_n=(\mathbb{Z}_m)^n\rtimes S_n$ generated by generalized prefix reversals. In this paper, we establish that, for all $m,n\geq2$, the spectral gap $\gamma(P(m,n))$ of the normalized Laplacian satisfies $\alpha_m/n\leq\gamma(P(m,n))\leq1/n$, where $\alpha_m$ is a positive constant that depends only on $m$. As a consequence, for every fixed $m\geq2$, $\gamma(P(m,n))$ is $\Theta_m(1/n)$ as $n\to\infty$. The proof combines Cesi's semi-recursive spectral-gap inequality with a Fourier decomposition of the appropriate operators associated with a coset Schreier graph of color-position pairs. For fixed $n\geq2$, we also establish that $\gamma(P(m,n))$ is $\Theta_n(m^{-2})$ as $m\to\infty$. This disproves a conjecture of Blanco and Buehrle asserting that, for fixed $n$, the corresponding undirected generalized pancake graphs form an expander family. Additionally, we present a counterexample to a recent conjecture of Greaves and Zhu concerning equality between the spectral gaps of the full Cayley graph and the associated coset Schreier graph.
Read the full abstract

Social and Information Networks

Consensus and Persistent Harmonic Edge Circulation in a Hodge-Theoretic Model of Networked Information Flow

We propose a finite-dimensional cochain model for information flow on an online communication complex. Node variables represent issue positions, while edge variables represent independently modelled signed information flow. The coupling is written in terms of the coboundary operator $d_0$ and the $1$-cochain Hodge Laplacian $\Delta_1=d_0d_0^*+d_1^*d_1$. We prove conservation of the mean opinion, a Lyapunov energy law, and convergence to an equilibrium determined by the initial harmonic projection of the edge flow. In particular, node opinions converge to consensus for every initial condition, whereas the edge flow converges to $P_{\mathcal H^1}u_0$. Thus, trivial first cohomology implies decay of the entire edge-flow variable. Nontrivial first cohomology supplies a topological capacity for residual edge circulation. A particular trajectory has a nonzero residual circulation if and only if $P_{\mathcal H^1}u_0\neq 0$, in which case the limiting edge flow is $P_{\mathcal H^1}u_0$. The model therefore establishes that node consensus need not imply decay of an independently represented edge-flow variable; it does not model the formation, reinforcement, or amplification of behavioral echo chambers. We also consider linear damping and a bounded nonlinear saturation as modifications that remove persistent edge flow under the stated assumptions.
Read the full abstract

Artificial Intelligence

Large Language Model Assisted Operational Monitoring for Battery Energy Storage System Integrated Power Distribution Networks

Battery energy storage systems (BESS) are increasingly used in distribution networks for voltage regulation and demand response, which increases the volume and complexity of operational telemetry available to grid operators. This paper presents an AI-enabled monitoring framework that connects a large language model (LLM) interface with a structured telemetry database for BESS-integrated distribution system analysis. Operator questions are submitted in natural language and translated into validated SQL queries using predefined database schema information and approved KPI views. Retrieved measurements, including bus voltages, state of charge, active power, and reactive power, are evaluated against engineering constraints for voltage limits, BESS operation, and demand response tracking. The framework is validated using hardware-in-the-loop co-simulation data from a BESS-equipped distribution feeder operating under reactive power-based voltage control and price-driven demand response. Case studies show that the framework generates valid database queries, identifies repeated voltage violations, detects reactive power overshoot, and evaluates active-power tracking performance. The results show that LLM-assisted monitoring can connect structured grid telemetry with automated engineering assessment for BESS operation analysis.
Read the full abstract

Computer Vision

JoLT: Joint Latent Trajectories for Context-Guided High-Resolution Tiled Generation

Although text-to-image generative models produce impressive results, they struggle to generate densely detailed, high-resolution (HR) images. Current literature addresses this issue with a low-to-high-resolution approach. First, a low-resolution (LR) image is generated. Then, an upsampled version is generated using the LR image as an additional cue. In this paper, we present Joint Latent Trajectories (JoLT). To generate an image, JoLT uses two streams that jointly denoise LR and HR latent images at each sampling step. The LR latent controls the overall layout, while the HR latent controls the details. We interconnect both branches to jointly integrate their information. We extensively validate our method, demonstrating its advantages over competing baselines. The resulting images are not only richly detailed but also visually pleasing, opening new avenues for artistic creation.
Read the full abstract

All authors Mathis Koroglu, Guillaume Jeanneret, Hugo Caselles-Dupré, Matthieu Cord, Arnaud Dapogny

Computation and Language

The Machine's Internal Clock: Do LLMs Share Human Temporal Illusions?

Human perception of time is subjective. Temporal illusions show that the brain relies on context and relational cues when judging duration. Prior studies established these effects with visual and auditory stimuli. We ask whether written narratives alone can cause these illusions, and whether LLMs respond as human participants do. We introduce a benchmark of 6,684 paired scenarios spanning five illusions, each judged from the character's and the reader's perspective. In our human study (60 participants), participants' judgments matched the predicted directions for all five illusions in at least one perspective (7 of 10 conditions). Across 14 LLMs, models reproduce every effect humans show in the same direction, but on average more strongly. Most effects persisted even after cue words were removed and after controlling for 15 lexical and semantic features. These results show that models and humans agree in direction but not in strength, which cautions against treating LLM judgments as substitutes for human responses for subjective temporal reasoning.
Read the full abstract

Artificial Intelligence

Visible Reasoning and Indirect Prompt-Injection Monitorability Across English, Tamil, and Tanglish

Chain-of-thought monitoring is a potentially useful safety signal, but its reliability across languages and behavioral settings remains uncertain. In a small case study of eight manually verified synthetic scenarios, one model, one annotator, and one deterministic generation seed, I study API-visible reasoning during indirect prompt injection in Sarvam-105B across English, Tamil, and Tanglish. A four scenario pilot found 5/12 injected attack successes without reasoning and 1/11 with reasoning. A preregistered four-scenario follow-up reversed that direction, finding 2/12 attacks without reasoning and 3/12 with reasoning. With only four scenarios per phase, this design cannot distinguish a real reasoning-mode effect from prompt-specific variation or sampling noise. Across 20 non-empty injected-thinking traces, all 17 benign-correct outputs stated an intent to ignore the injection, while all three attack successes stated an intent to follow it. These descriptive observations provide a reproducible case study of behaviorally informative visible reasoning when it is available; they do not establish that reasoning mode improves safety, that visible reasoning is mechanistically faithful, or that the findings generalize beyond this configuration.
Read the full abstract

Artificial Intelligence

TwinGridShield: Consequence-Aware Runtime Authorization for LLM Grid-Agent Actions

Large language model (LLM)-assisted energy-management tools can translate natural-language context into structured grid commands, but syntactic validity does not imply physical admissibility. This paper presents TwinGridShield, a model-independent runtime authorization layer that evaluates each proposed action in a deterministic network twin before release. The prototype checks connectivity, branch-flow, generator, and load-shedding invariants and records each decision in a hash-chained log. A controlled IEEE 14-bus study evaluates single-step switching, redispatch, and load-shedding actions using DC power flow and experimentally assigned branch ratings. In the matched-model experiment, a stochastic proposal source configured to select an unsafe action with probability p=0.84 produced 421 unsafe proposals in 500 attacked-condition trials, a realized rate of 84.2%. This value characterizes the configured surrogate and is not an empirical measurement of LLM prompt-injection susceptibility. TwinGridShield produced 0 unsafe releases in those 500 trials. Because action labeling and authorization used the same DC model, system state, branch ratings, and encoded constraints, this result verifies conformance of the implementation to its encoded authorization predicate rather than safety under model error. The principal robustness evaluation therefore introduces model mismatch. Unsafe acceptance reached 5.63% under bounded +20% and -20% per-bus load-measurement error and 30.09% when actual branch ratings were 20% below modeled ratings.
Read the full abstract

Data Structures and Algorithms

The Quick and the Dead: Estimating Sparse-Matrix Permanents with Adaptive Work Filtering

Rasmussen's permanent estimator is a simple and unbiased estimator for the permanent of a binary matrix, but its practical performance can be limited by trajectories that terminate before completing a perfect matching. These failed trajectories, together with dispersion among the surviving weights, can substantially reduce the effective sample size. Although the literature leverages techniques such as matrix scaling to improve proposal balance and support filtering to remove structurally infeasible choices, using these at every step can substantially increase the trajectory cost. Furthermore, they do not directly address the choice of the next vertex. This paper uses the classical minimum-degree ordering in sparse matrix algorithms to select the next vertex with O(n + m) total bucket-maintenance work per trajectory, where n is the number of rows/columns in the matrix and m is the number of nonzeros. The proposed estimator uses adaptive schedules to invoke the more expensive scaling and filtering operations only when needed. The experiments show that it is competitive with the state of the art on the tested small matrices and scales effectively to large sparse matrices.
Read the full abstract

All authors Deniz Elbek, Yiğit Manafi, Zeynep Gürdal, Sinan Y{\i}ld{\i}r{\i}m, Kamer Kaya

Artificial Intelligence

Agentic-SQL Revisited: Autonomy-Based Taxonomy and Empirical Benchmark Analysis for LLM Text-to-SQL

LLM-based Text-to-SQL progress is reported across heterogeneous benchmarks, backbones, and inference protocols, making cross-system comparison fragile. We reframe the field as a leaderboard aggregation: we collect the metrics authors themselves report and organize them along an inference-autonomy axis spanning constrained, in-context, iterative, agentic, and reasoning-internalized generation, with traceable provenance for every cell. To anchor the aggregation empirically, we run a focused case study on Spider, comparing 8B open-source backbones with and without chain-of-thought (CoT) supervision against few-shot DeepSeek~V3 and GLM-4 baselines. Four patterns emerge: Spider gains transfer unevenly to BIRD and Spider~2.0; autonomy buys robustness at non-trivial cost; reasoning internalization sits between answer-only decoding and externally orchestrated agents; and CoT gains concentrate on Hard and Extra-Hard queries. We release a Python harness mirroring the autonomy axis so that future methods can be added directly to the leaderboard.
Read the full abstract

All authors Yiyun Su, Zujun Peng, Yu Tian, Yuting Liu, Changruo Zhao, Huiying Zhu, Luyan Zhang, Heming Zeng

Machine Learning

Look Before You Lift: Visual and Quantitative Diagnostics for Topological Deep Learning

Topological deep learning (TDL) methods rely on lifting raw data into higher-order discrete domains such as simplicial complexes, cell complexes, and hypergraphs. In practice, this lifting step is often treated as a black box: practitioners select a lifting and then tune architectures, with limited visibility into whether the induced higher-order connectivity is meaningful for the downstream task. To address this missing diagnostic layer, we propose a visualization technique called TopoExplorer that leverages the strictly augmented Hasse graph form of topological datasets for exploratory data analysis. For the first time, practitioners can easily visualize the incidence- and adjacency-based neighborhoods that define the lifted dataset, as well as read off key graph metrics that describe its structural and feature landscape. Via an extensive set of experiments across many datasets and liftings, we show that several of these metrics correlate with downstream model performance, suggesting they can help inform TDL preprocessing design. Our perspective reframes the TDL workflow from lift-train to lift-look-design-train, enabling more principled, interpretable, and efficient model development. TopoExplorer is hosted at https://topoexplorer.pagekite.me, and its source code is available at github.com/geometric-intelligence/topoexplorer.
Read the full abstract

All authors Mathilde Papillon, Guillermo Bernárdez, Álvaro Ballón Barreiro, Marco Montagna, Rémi Devaux, Antoine Jardin, Nina Miolane

Software Engineering

On the Influence of Refactoring Types on Merge Effort

Modern software development involves parallel work and concurrent changes, requiring code merging. Prior studies report that 10% to 20% of merge attempts result in conflicts, often requiring manual intervention. The literature explores factors that generate conflicts, including refactorings, but does not analyze how individual refactoring types influence the manual effort required to resolve them. We analyzed 64 open-source Java projects and applied association rule mining to measure the strength of associations between specific refactoring types and merge effort. Our results show that refactoring types relate to merge effort with varying strength. In particular, Rename Attribute, Move Class, Extract Variable, Change Return Type, and Split Parameter exhibit some of the strongest associations, especially when a higher number of such refactorings is present in the merge branches. We also find that both the number of refactorings and their diversity independently increase merge effort, both in terms of occurrence and intensity. Additionally, the co-occurrence of refactorings across parallel branches is associated with higher merge effort, particularly when combining structural transformations with changes to method signatures and data-structure representations, whereas more localized changes are less frequent in the most impactful combinations.
Read the full abstract

All authors André Oliveira, João Victor Monteiro, Vânia Neves, Alexandre Plastino, Bianca Trinkenreich, Alessandro Garcia, Leonardo Murta

Machine Learning

Every Expert Counts: ExactMoE for Memory-Efficient W4A16 Inference

Sparse mixture-of-experts (MoE) language models reduce arithmetic by activating only a small subset of experts per token, yet deployment still requires storing and moving the full expert bank. We present ExactMoE, an inference design that applies symmetric group-128 four-bit weight quantization only to routed experts, stores those experts in kernel-native MARLIN form in pinned host memory, and executes all selected experts through a configurable GPU-resident slot cache and fused grouped MoE kernels. The router, attention, embeddings, normalization layers, and language-model head remain in BF16. "Exact" refers to complete expert availability and an unchanged top-k routing procedure: no expert is pruned, substituted, or forced to execute on the CPU. It does not imply numerical identity with the BF16 model. On OLMoE-1B-7B-0924-Instruct, evaluated on a single NVIDIA L4, a 16-slot configuration reduces peak reserved GPU memory from 14.168 to 1.836 GiB (87.04%) while retaining 81.85% of BF16 decode throughput. A fully resident 64-slot configuration reaches 31.923 tokens/s versus 21.662 tokens/s for BF16 while reserving 4.061 GiB. Across 12,450 zero-shot multiple-choice questions, ExactMoE obtains 70.3534% normalized accuracy versus 70.8996% for BF16, retaining 99.23% of the baseline accuracy. In a matched 16-token ablation, fused grouped execution is 1.97x as fast as a sequential W4 reference. These results identify a practical memory-transfer-throughput frontier for complete-expert MoE inference.
Read the full abstract

Artificial Intelligence

Framework for Grounding Healthcare LLMs in a Causal Knowledge Graph: A Cardiovascular Example Pilot

Large language models (LLMs) are increasingly proposed for healthcare decision support, but their evaluations still reward single-answer accuracy rather than reasoning about interventions, mechanisms, harms, evidence, and uncertainty. We propose a reproducible, graph-centered evaluation framework for intervention-oriented LLM behavior in healthcare and stress-test it in a cardiovascular pilot. The framework has four components: (i) a domain causal knowledge graph in which assertions are first-class, provenance-preserving nodes with stable identifiers; (ii) a scenario-conditioned subgraph extraction step that, given any clinical scenario, retrieves the relevant reified-assertion subgraph; (iii) four controlled grounding conditions that vary how the retrieved subgraph is composed into the model's context (ungrounded C1, knowledge-graph C2, causal-graph C3, integrated C4); and (iv) an automated scoring pipeline, anchored on assertion identifiers, that computes intervention accuracy, and other evaluation measures on a single pass. To test the framework, we built a category-balanced scenario generator across eight reasoning failure modes and instantiated it on a cardiovascular graph. The metric panel discriminates conditions along interpretable, non-redundant axes: C4 obtains the strongest causal edge F1 (0.838), adverse-effect F1 (0.833), evidence accuracy (0.738), and unsupported claim rate (0.114), while C1 obtains the highest raw intervention accuracy (0.948) with no measurable causal or evidential grounding.
Read the full abstract

Artificial Intelligence

FedPA-LoRA: Product-Aligned Framework for Mitigating Aggregation and Initialization Errors in Heterogeneous Federated LoRA

Low-Rank Adaptation (LoRA) enables efficient federated fine-tuning of large language models, but its factorization creates a tension between aggregation fidelity and local optimization continuity. Factor-wise aggregation incurs aggregation mismatch but better preserves factor continuity, whereas product-space reconstruction reduces this mismatch at the cost of greater factor-level initialization mismatch. We propose FedPA-LoRA, a product-aligned framework that addresses both mismatches and establishes convergence bounds for homogeneous and heterogeneous ranks. Clients preserve local factors across rounds to eliminate initialization mismatch and align their products with rank-specific global references to reduce client drift. The server aggregates heterogeneous-rank updates in product space to mitigate aggregation mismatch and efficiently reconstructs a rank-constrained global adapter without forming the dense aggregate. By decoupling local and reference ranks, FedPA-LoRA supports client-specific computation and communication budgets. Experiments on natural language understanding and generation tasks show consistent gains under data and resource heterogeneity, including a 6.82 percentage-point improvement in average GLUE accuracy under heterogeneous ranks. The code is available at https://github.com/Juseok-Jeon/FedPA-LoRA.
Read the full abstract

All authors Juseok Jeon, Ramy E. Ali, Doyun Kwon, Myungbeom Her, Jinhwi Kim, Jinhyun So