Artificial Intelligence
Neurosymbolic Alignment for Physiologically-Safe Clinical Language Models
Abdulhady Abas Abdullah, Erik Cambria and Milena Zivkovic
Clinical LLMs can generate recommendations that are factually plausible yet physiologically unsafe. We investigate whether safety
alignment can be improved by grounding preference optimization in structured physiological knowledge rather than text-only supervision. Methods: We propose Neurosymbolic
Alignment, a training-time framework that couples a 7B clinical LLM with an HGNN-based Physiological World Model over an 847K-node biomedical knowledge graph. Candidate responses are scored using homeostatic constraints, multi-hop path plausibility, and drug-interaction penalties, and the resulting rankings drive iterative on-policy ORPO updates. Evaluation is performed on the Clinical Safety
Benchmark (CSB), a 2,500-scenario
benchmark for physiological constraint violations in generative clinical
reasoning. Results: Relative to ORPO, the proposed method improves CSS from 69.5% to 90.8% (+21.3 pp), reduces physician-evaluated HR from 14.1% to 5.1% on the blinded subset, and improves DID from 72.8% to 91.6%. These gains are corroborated by an HGNN-independent Rule-Engine Safety Score (RSS: 86.4%, +21.2 pp over ORPO; r=0.97 concordance with CSS). The method also exceeds GPT-4 (5-shot) on all safety metrics despite a 10x parameter disadvantage, and outperforms an inference-time self-correction pipeline (SFT+SelfCorrect) by 11.4 pp CSS. Under synthetic EHR-style noise, 84.2% CSS is retained. Ablation analysis shows that HGNN scoring (-16.2 pp) and iterative training (-11.5 pp) are the dominant contributors. PhysioScore calibration against 200 clinician labels yielded ECE = 0.038 and kappa = 0.91. Conclusion: Training-time physiological grounding produces measurable and independently verifiable safety improvements in open-weight clinical LLMs under controlled evaluation. External validation on real clinical data is required to determine whether these gains transfer to deployment settings
Read the full abstract ▾
Computational Engineering
MoRF-AST: Calibrated Probabilistic Virtual Sensing for Structural Monitoring under Changing Operating Conditions
Wingho Feng, Quanwang Li, Ming Zhong et al. (+2 more)
Probabilistic full-field reconstruction provides uncertainty-aware response evidence for structural reliability assessment, yet inference from sparse and noisy measurements remains underdetermined. Most existing methods overlook shifts between offline training and operational distributions. Under such shifts, posterior intervals may become miscalibrated, causing the reported uncertainty to lose its probabilistic meaning. This study proposes Modal Residual Flow Matching with Context-Conditioned Affine Spread Transport (MoRF-AST) for calibrated structural virtual sensing under changing operating conditions. MoRF constructs an analytic Gaussian reference posterior in normalized modal coordinates and trains a conditional flow only on posterior-whitened residuals. At deployment, AST estimates response scale from historical measurements at installed sensors and uses gated, mean-preserving Bures-Wasserstein transport to adjust posterior spread. On a bridge-deck
benchmark, MoRF achieves a posterior-mean normalized root-mean-square error (NRMSE) of 7.20%, compared with 16.1% and 17.9% for two direct conditional flows. Across eight shifted traffic domains, AST reduces MoRF's cross-domain average coverage error from 0.0535 to 0.0236, a 55.9% reduction, while preserving posterior-mean accuracy. The same transport does not improve the tested alternatives in aggregate, showing that calibration gains require its direction to match the base posterior's dispersion bias. MoRF-AST provides a data-efficient framework for probabilistic full-field reconstruction whose uncertainty remains interpretable under scale-dominated operational distribution shifts. More broadly, this work highlights the need to calibrate uncertainty under changing operational distributions, thereby supporting trustworthy probabilistic modeling and reliability-informed decision-making in civil and infrastructure engineering.
Read the full abstract ▾
All authors Wingho Feng, Quanwang Li, Ming Zhong, Jingyu Yang, Chen Wang
Numerical Analysis
Randomized Block Davidson Eigensolvers for Plane-Wave Density-Functional Theory
Moritz Gubler, Taejun Park, Augustin Bussy et al. (+4 more)
Iterative diagonalization is the dominant cost of plane-wave density-functional theory (DFT), with search-space orthogonalization scaling particularly quickly with problem size and the number of target states. We present a randomized block Davidson-type eigensolver that replaces Euclidean orthogonalization with randomized Gram-Schmidt in a sketched inner product, requiring only a single pass over the basis while keeping its conditioning bounded independently of the input vectors. This modification changes only the Rayleigh-Ritz step, which becomes a definite generalized Hermitian eigenproblem. Ritz extraction remains exact, preserving true Ritz pairs and the interlacing property that makes each band energy an upper bound on the true one. The method is implemented in mixed precision for CPUs and GPUs from a single Julia code, interfaces
matrix-free with DFTK, and is released in the open-source RandESC library. On sparse test problems with a fixed number of eigenpairs, the sketched solver overtakes its deterministic counterpart beyond
matrix dimensions of about $2\times 10^4$ and is $25\%$ faster at $5\times 10^5$. In full self-consistent field DFT calculations, however, both Davidson variants outperform the locally optimal block preconditioned conjugate gradient (LOBPCG) reference only by $5$ to $11\%$ in total time, while the additional benefit of sketching is limited. As the number of requested states grows with system size, orthogonalization savings are offset by the generalized eigenproblem. Therefore, the regime in which sketching pays off is set by how the number of wanted states scales with the problem dimension, not by the eigensolver as such.
Read the full abstract ▾
All authors Moritz Gubler, Taejun Park, Augustin Bussy, Johannes Laute, Nicola Marzari, Iurii Timrov, Laura Grigori
Data Structures and Algorithms · Cross-listed from quant-ph
Provable Quantum-Classical Separation for Continuous Gibbs Sampling
Enrico Olivucci, Mariia Sobchuk, Sehmimul Hoque et al. (+4 more)
We prove the first
quantum-classical separation for a sampling problem over a continuous domain. For a class of Gibbs states $p\propto e^{-\beta E}$ on the
torus $\mathbb{T}^d$ with smooth ($s$-Gevrey) potential and barrier amplitude $\alpha=e^{\beta\Delta}$, where $\Delta = \max E-\min E$, every classical
algorithm querying the value, gradient, or any higher-order derivatives of the log-density requires $\Omega(\alpha)$ queries to sample at constant accuracy in total variation distance, while a
quantum algorithm based on
quantum singular value thresholding and temperature annealing samples with $\tilde{O}\left(\sqrt{\alpha}\right)$ queries to an oracle for the gradient. The advantage is quadratic in the barrier amplitude, which becomes exponential in the dimension, $e^{\Omega(d)}$, at low temperature. The classical bound is information-theoretic, holding for every classical
algorithm with query access to the Gibbs potential and its derivatives at any order.
Read the full abstract ▾
All authors Enrico Olivucci, Mariia Sobchuk, Sehmimul Hoque, Jeffrey Hnybida, Kyungho W. Kim, Ala Shayeghi, Pooya Ronagh
Robotics
RoG-DAgger: Rollout-Guided Post-Training for End-to-End Driving
Liangyu Zhong, Joachim Sicking, Fabian Hueger et al. (+1 more)
Recent end-to-end driving systems demonstrate strong performance on closed-loop
benchmarks, yet are still predominantly trained on fixed expert-collected data using open-loop imitation learning. This training-inference mismatch leaves the policy vulnerable in policy-induced states, where accumulated errors can lead to safety-critical failures. A promising post-training approach to overcome this issue is
Dataset Aggregation (DAgger), which gathers expert demonstrations in policy-induced states and subsequently fine-tunes the policy on the resulting aggregated
dataset. Existing driving DAgger pipelines, however, face three challenges: i) the expert is restricted to a limited trajectory-and-speed solution space, ii) takeover may occur too early or too late relative to impending failures, and iii) privileged expert decisions may rely on information unavailable to the student. To address this, we introduce RoG-DAgger, a post-training framework that uses short-horizon kinematic rollouts to construct high-quality expert demonstrations in safety-critical states. Specifically, RoG-DAgger expands the expert's trajectory-and-speed solution space and evaluates candidate plans through rollout to construct preventive supervision. Moreover, it uses rollout solvability to time the takeover near the estimated point of no return. Lastly, it aligns the expert's field of view with that of the student to provide student-compatible supervision. Across in-distribution (including long-horizon) and out-of-distribution evaluations, RoG-DAgger improves the end-to-end model SimLingo by 5.3 driving-score points and 6.2 percentage points in success rate on Bench2Drive, doubles its driving score from 22 to 44 on Longest6 v2, and improves out-of-distribution success rate from 55% to 66% on Fail2Drive.
Read the full abstract ▾
All authors Liangyu Zhong, Joachim Sicking, Fabian Hueger, Hanno Gottschalk
Software Engineering
LumiXAI: A Modular Full-Stack Framework for Feature Attribution
Alfio Ferrara, Lorenzo Gatta, Sergio Picascia et al. (+1 more)
Feature attribution is a central tool of model interpretability, yet the software through which it is applied remains fragmented: individual tools specialize along narrow axes, such as a single modality, a code API or a GUI, or a fixed rather than extensible method set, and rarely combine these strengths. Moreover, many explainability tools are designed primarily for domain experts, requiring programming skills or familiarity with attribution methods that can make them difficult for non-expert users to access. In this article, we present LumiXAI, a modular full-stack framework that consolidates attribution analysis into a single system. It couples classification and generative attribution with an interactive GUI supporting bidirectional exploration, a plug-in architecture for registering new models and methods, and three access tiers serving non-programmers, developers, and extenders from one backend. Its contribution is a system that operationalises established attribution methods under one interface, one interaction model, and one persistence layer, with containerised services and persistent results making analyses reproducible across machines.
Read the full abstract ▾
All authors Alfio Ferrara, Lorenzo Gatta, Sergio Picascia, Elisabetta Rocchetti
Computation and Language
Beyond Information Seeking: Severity-Aware Question Supervision for Proactive Medical Dialogue
Chenxuan Li, Xinrong Chen, Luyan Zhang et al. (+5 more)
Proactive medical dialogue requires an
agent to decide what to ask from incomplete patient information. Existing information-seeking approaches commonly prioritize questions that most reduce diagnostic uncertainty, but this criterion overlooks an important property of medical diagnosis: different diagnostic errors can carry substantially different consequences. The most informative question may therefore differ from the one most valuable for the downstream decision. We propose Expected-Severity-Risk (ESR), a consequence-aware question-supervision objective that values each candidate by its expected reduction in severity-aware terminal risk. Because questions must be selected before their answers are observed, ESR marginalizes over possible answers using train-only population statistics. Its rankings are then distilled into a prefix-only
language policy, requiring no teacher-side risk computation at deployment. Across three matched Qwen3-4B training seeds on DDxPlus, ESR reduces mean high-severity diagnostic miss from 0.0645 to 0.0455 (29.5% relative reduction) and improves mean diagnostic accuracy from 0.9123 to 0.9320 while requiring only 0.14 additional questions per dialogue. Fixed-budget analyses show that the distinction persists when question count is controlled, while a matched expected-0/1-risk student control further isolates the contribution of asymmetric severity weighting. These results support moving proactive medical dialogue beyond uncertainty reduction toward consequence-aware evidence acquisition.
Read the full abstract ▾
All authors Chenxuan Li, Xinrong Chen, Luyan Zhang, Peidong Jia, Runfan Zheng, Zhongyu Zhao, Xuecheng Shang, Peixing Wan
Computational Complexity · Cross-listed from math.CO
Homomorphic-core phase transition threshold in Erdős–Rényi random graphs
Jiaheng Wang
It is shown in this manuscript that a random graph $G$ drawn from the Erdős–Rényi model $\mathcal{G}(n,p)$ with \[ p=p(n)\leq 1/2, \qquad \lim_{n\to+\infty}(np-\log n-\log\log n)=+\infty, \] is a homomorphic core, i.e., every homomorphism from $G$ to itself is an
automorphism. This implies tight ETH-based lower bounds of the subgraph isomorphism problem for almost all $k$-vertex patterns with
polynomial average degree.
Read the full abstract ▾
Machine Learning
It depends: Incorporating correlations for joint aleatoric and epistemic uncertainties of high-dimensional output spaces
Leonhard F. Feiner, Manuel Nickel, Martin Menten et al. (+6 more)
Uncertainty Quantification (UQ) plays a vital role in enhancing the reliability of deep learning model predictions, especially in scenarios with high-dimensional output spaces. This paper addresses the dual nature of uncertainty – aleatoric and epistemic – focusing on their joint integration in high-dimensional regression tasks. For example, in applications like medical image
segmentation or restoration, aleatoric uncertainty captures inherent data noise, while epistemic uncertainty quantifies the model's confidence in unfamiliar conditions. Modeling both jointly enables more reliable predictions by reflecting both unavoidable variability and knowledge gaps, whereas modeling only one limits transparency and robustness. We propose a novel approach that approximates the resulting joint uncertainty using a low-rank plus diagonal covariance structure, capturing essential output correlations while avoiding the computational burdens of full covariance
matrices. Unlike prior work, our method explicitly combines aleatoric and epistemic uncertainties into a unified second-order distribution that supports robust downstream analyses like sampling and log-likelihood evaluation. We further introduce stabilization strategies for efficient training and inference, achieving superior UQ in the tasks of image inpainting, colorization, optical flow, and depth estimation.
Read the full abstract ▾
All authors Leonhard F. Feiner, Manuel Nickel, Martin Menten, Laurin Lux, Rickmer Braren, Daniel Rueckert, Georgios Kaissis, Raphael Rehms, Johannes Paetzold
Distributed Computing
SatDL: Jointly Optimizing Data Redistribution and Training for Satellite-Based Distributed Learning
Hao Wu, Kin Whye Chew, Yizhan Han et al. (+2 more)
Satellite-based distributed learning promises to train machine-learning models directly in orbit using massive, globally dispersed sensor data, thereby avoiding large-scale data downloads to ground servers. However, training convergence is significantly slowed by severe non-IID data, specifically label imbalance, as each satellite observes different geographic regions with distinct labels. This imbalance extends training duration and increases energy consumption for
solar-powered satellites. Existing approaches either fully redistribute data to enforce IID conditions - accelerating convergence but incurring substantial communication delays - or avoid redistribution entirely by modifying local learning
algorithms to mitigate the impact of label imbalance, which, however, still prolong training and increase energy use. Both extremes result in excessive total end-to-end learning time (data-transfer delay plus training time) and thus elevated onboard energy consumption. We present SatDL, a data-redistribution framework designed to minimize total end-to-end learning time. At its core, SatDL develops a Distributor-Critic framework that jointly models and optimizes data-transfer delay and training time. Evaluations through trace-driven simulations of a 1,584-satellite Starlink constellation and hardware emulations using NVIDIA Jetson and A100 GPUs across five
datasets show SatDL reduces total end-to-end learning time by up to 18.6% and onboard energy consumption by 12.23-88.00%, while maintaining inference accuracy within a few percentage points of state-of-the-art baselines.
Read the full abstract ▾
All authors Hao Wu, Kin Whye Chew, Yizhan Han, Han Li, Jingxian Wang
Data Structures and Algorithms
Designing Caterpillars for Graphs: Approximation and Hardness
Leon Kullmann, Phuoc Lucky Trinh, Leon Kellerhals et al. (+3 more)
The classical Minimum Linear Arrangement (MLA) problem has been studied extensively. It is known to be NP-hard and it admits an $O(\sqrt{\log n}\log\log n)$-approximation [Feige and Lee, IPL, 2007]. MLA can be defined as follows as design problem: Given a graph $G$ with vertex set $V(G)$, design a path $H$ on the same vertex set that minimizes the linear arrangement cost $\sum_{uv\in E(G)}\textrm{dist}_H(u,v)$, where $\textrm{dist}_H(u,v)$ indicates the distance of $u$ and $v$ in $H$. We initiate the study of the generalization in which $H$ is allowed to be a caterpillar graph of maximum degree at most $\Delta$. Caterpillars are the simplest generalization of paths, having pathwidth one and interpolating between paths and
stars via the degree parameter $\Delta$. We give an
algorithm that lifts any $\alpha$-approximation for MLA to an $(\alpha+3-2/(\Delta-1))$-approximation for our problem, thus obtaining an $O(\sqrt{\log n}\log\log n)$-approximation for our more general problem as well. Moreover, we derive a $4$-approximation whenever MLA is
polynomial-time solvable, in particular, for trees. Complementing these results, we prove NP-hardness for every constant $\Delta\geq 2$, and, in stark contrast to MLA, show it remains NP-hard on trees when $\Delta$ is part of the input.
Read the full abstract ▾
All authors Leon Kullmann, Phuoc Lucky Trinh, Leon Kellerhals, Mitja Krebs, André Nichterlein, Stefan Schmid
Artificial Intelligence
PeakBench: Benchmarking Resource-Aware Tool Invocation in LLM Agents
Zhi-Kai Chen, Xu-Xiang Zhong, Song-Yan Li et al. (+2 more)
LLM
agents increasingly solve tasks by invoking multiple tools, where parallel execution is essential for low latency but difficult to manage safely. Existing
agent benchmarks primarily evaluate tool selection, argument generation, and end-to-end success under mostly serial execution, largely overlooking valid parallelization and resource-constrained scheduling. This missing scheduling dimension creates a practical failure mode: serial execution is safe but slow, while resource-agnostic parallel execution is fast but prone to avoidable resource overflows. To address this gap, we introduce PeakBench, a
benchmark of executable multi-tool workflows with execution-grounded dependency annotations and measured resource profiles. A central challenge in evaluating such workflows is attribution: failures and inefficiencies may arise from incorrect dependency planning, poor resource-constrained scheduling, or both. PeakBench addresses this challenge with a two-part evaluation framework that disentangles logical planning from physical scheduling, with dedicated metrics for each dimension. Using this framework, we show that strong logical planning does not reliably translate into safe or efficient execution under resource constraints. We further show that exposing resource information can reduce avoidable overflows and improve resource utilization, making PeakBench a useful testbed for diagnosing resource-aware
agent behavior. Code is available at https://github.com/Czzzk/Staggering-the-Peaks.
Read the full abstract ▾
All authors Zhi-Kai Chen, Xu-Xiang Zhong, Song-Yan Li, De-Chuan Zhan, Han-Jia Ye
Information Theory
Achieving Torn-Paper Channel Capacity with Successive Revelation
Rui Xu and Le Wang
The torn-paper channel independently cuts a binary codeword at its internal boundaries and outputs the resulting oriented fragments as an unordered multiset. We consider the critical regime pN log N to alpha, in which the channel capacity is e to alpha. Existing coding schemes use a fixed-density pilot to localize fragments, creating a tradeoff between positional information and payload rate. This paper introduces successive revelation, which partitions the codeword into interleaved tracks and decodes them sequentially. Each recovered payload track becomes an additional positional reference for subsequent stages, allowing progressively shorter fragments to be localized. We establish a finite-track achievable rate whose gap to capacity is O 1 or M for M tracks. Consequently, for every rate below the channel capacity, a finite number of tracks yields a sequence of deterministic codes with vanishing average decoding error probability.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from stat.ML
Scalable and Versatile Identification for Hierarchical Structural Causal Models: A New Look at Project STAR
Janis Aiad, Aghiles Drali, Aymen El Ouadrhiri et al. (+6 more)
The
STAR (Student-Teacher Achievement Ratio) experiment (1985, Tennessee, USA) is a landmark hierarchical
dataset designed to assess the impact of class size on student outcomes, with observations nested within classes. To encode class-level interventions in such hierarchical settings, we develop a complete, scalable, open-source pipeline for Hierarchical Structural Causal Models (HSCM) that bridges symbolic identification and practical estimation. Our approach integrates graph transformations, pyAgrum's do-calculus for automatic identification of causal effects, adaptation of symbolic expression into closed-form HSCM formulas, and numerical estimation from fitted local probability models. A key innovation is our adapted Abstract Syntax Tree (AST), which decomposes pyAgrum's identified formulas into independent density, expectation, and marginalization tasks, enabling parallel and scalable computation. We validate the pipeline on canonical HSCM motifs and
benchmark scenarios with known ground truth, then apply it to
STAR kindergarten mathematics outcomes. The results show that flat baselines (ignoring hierarchy) recover associations but fail to encode class-level interventions, and that symbolic identification alone is not enough for practical Hierarchical Structural Causal inference; scalable estimation and numerical stability checks are central parts of the scientific object.
Read the full abstract ▾
All authors Janis Aiad, Aghiles Drali, Aymen El Ouadrhiri, Anass Ettahiri, Yasser Oufqir, Simon Patry, David Cortes, Marianne Clausel, Emilie Devijver
Cryptography and Security
SeriCrypt: An LLM-Driven Context-Aware Serialization Framework for Cryptographic Protocols
Maosong Chen, Xi Chen, Mengcheng Ju et al. (+2 more)
Constructing syntactically correct and cryptographically valid message sequences is essential for protocol state machine learning, conformance testing, and fuzzing. Unlike plaintext protocols, cryptographic protocols involve complex cross-message state dependencies and cryptographic computation constraints. Existing automated approaches predominantly target text-based or plaintext protocols, leaving cryptographic message construction largely manual. We present SeriCrypt, an LLM-driven, context-aware serialization framework for cryptographic protocols. It employs a large
language model to extract field constraints, state dependencies, and cryptographic computation rules from unstructured protocol specifications into a unified structured intermediate representation, formally characterized by a domain-specific
language for cryptographic protocols (CDSL). A protocol-agnostic execution engine parses CDSL declarations, automating field value resolution, cryptographic primitive invocation, and byte-stream serialization. As case studies in protocol security testing, we use the framework to construct violation messages targeting specification-defined security constraints and to support protocol fuzzing, evaluating it on mainstream implementations of TLS 1.2/1.3, IKEv1/v2, SSH, and TLCP. SeriCrypt generated message sequences accepted by all evaluated implementations and completed handshakes in every scenario. Security constraint testing revealed five specification violations, and fuzzing reached deeper protocol states with higher code coverage than mainstream fuzzers under the same time budget, demonstrating the framework's practical value for cryptographic protocol security testing.
Read the full abstract ▾
All authors Maosong Chen, Xi Chen, Mengcheng Ju, Dongliang Zhao, Chunxiang Gu
Computational Geometry
Computing an e-net of a closed hyperbolic surface
Vincent Delecroix, Vincent Despré, Camille Lanuel et al. (+2 more)
Hyperbolic surfaces are a fundamental object in mathematics and play an increasingly important role in computational geometry and
topology. A key ingredient in the design of efficient
algorithms on such surfaces is the availability of a
geometric discretization of controlled complexity. In this paper, we present the first
algorithm for constructing e-nets on
hyperbolic surfaces starting from a fundamental polygon representation. Our approach is based on Delaunay refinement and relies on maintaining Delaunay triangulations through edge flips. The size of an e-net cannot be bounded solely as a function of the genus because of the presence of arbitrarily long collars around short geodesics. To overcome this difficulty, we introduce the notion of a pseudo e-net, which decomposes the surface into e-thin cylinders together with a Delaunay triangulation over an e-net of the remaining thick part. As applications, we obtain
algorithms for computing the length spectrum of an e-thick
hyperbolic surface and for computing the systole from a pseudo log(sqrt(2))-net. These results demonstrate that Delaunay-based discretizations provide a practical and versatile framework for algorithmic computations on
hyperbolic surfaces.
Read the full abstract ▾
All authors Vincent Delecroix, Vincent Despré, Camille Lanuel, Hugo Parlier, Monique Teillaud
Networking
A Capability Broker for Workflow-Network QoS Coordination in B5G/6G Industrial Services
Qize Guo, Yan Chen, Taleb Tarik et al. (+3 more)
Industrial services in beyond-fifth-generation (B5G) and sixth-generation (6G) networks are increasingly executed as multi-phase workflows whose quality-of-service (QoS) demands change across ordered phases. Existing exposure, analytics, and policy-control functions can provide capability information and enforce QoS treatment, but they do not track the admitted QoS agreement between a workflow and the network. This paper studies this missing coordination function in the software control plane. We propose a Capability Broker that represents each admitted trajectory as a lifecycle-managed QoS commitment, including the workflow demand, network capability, validity time, and enforcement state. The Broker operates above existing exposure and policy-control interfaces and enforces capability freshness, duplicate-safe admission, legal state transitions, per-commitment event ordering, recovery routing, and Broker–network state consistency. Prototype results with microbenchmarks, fault injection, ablation, and controlled message loss show that the full Broker preserves commitment consistency with microsecond-level local processing overhead, while removing individual Broker functions exposes duplicate commitments, stale admissions, illegal transitions, lost-update divergences, or Broker-network inconsistency.
Read the full abstract ▾
All authors Qize Guo, Yan Chen, Taleb Tarik, Bjoern Riemer, Hemant Zope, Hao Yu
Computational Complexity · Cross-listed from quant-ph
Optimal Quantum Algorithm for Ground-State Energy Estimation with a Guiding State
Stacey Jeffery and Freek Witteveen
In the problem of ground-state energy estimation, one aims to estimate the smallest
eigenvalue of a
Hamiltonian, often given a guiding state, with some promised overlap $\gamma$ with the ground space. The main approach to this problem is to simulate its
evolution, and estimate the smallest (or equivalently, largest) eigenphase of the resulting
unitary $U$. We give a
quantum algorithm that estimates the largest eigenphase of a
unitary $U$ in this guided setting using a factor of $\log\frac{1}{\gamma}$ fewer queries to $U$ than the previous best approach. The result matches an existing lower bound, and answers an open question from Mande and de Wolf. The
algorithm is based on transducers, which often allow composition of
quantum algorithms without overhead from error reduction.
Read the full abstract ▾
Computational Complexity · Cross-listed from quant-ph
Optimal Lower Bound for Ground-State Energy Estimation with a Guiding State
Rolando D. Somma and Ronald de Wolf
Suppose we can apply the
unitary $U=e^{i H}$ for some
Hamiltonian $H$, and are given access to a
unitary that prepares a guiding state promised to have overlap at least $\gamma >0$ with the ground space of $H$. Our goal is to estimate the ground-state energy of $H$ within additive error $\delta > 0$ and success probability at least $1-\varepsilon$, $\varepsilon>0$. How many applications of $U$ and its inverse $U^{-1}$ are necessary and sufficient? This quantity corresponds to the total
Hamiltonian-simulation time needed. An upper bound $O(\log(1/\varepsilon)\log(1/\gamma)/\gamma\delta)$ was known, and was improved to $O(\log(1/\varepsilon)/\gamma\delta)$ very recently [JW26]. A matching lower bound was known whenever one of the three parameters $\delta,\gamma,\varepsilon$ was held constant [MdW26]. In this paper we prove the joint lower bound $\Omega(\log(1/\varepsilon)/\gamma\delta)$ with the tight $\varepsilon$-dependence provided the dimension of $H$ is at least $\log(1/\varepsilon)/\gamma^2$. Furthermore, we show that this same lower bound (with slightly larger dimension) holds for both the special case in which the ground state is guaranteed to be unique and $H$ has a gap of $\delta$ between its first and second
eigenvalue; and for ground-state preparation, where $\delta$ denotes the spectral gap and $\varepsilon$ now is the approximation error. The lower bounds also apply when the
Hamiltonian can be accessed via its block-encoding, and when fractional powers of $U$ are allowed, as in continuous-time
Hamiltonian simulation. Lastly, improved upper bounds are known when $H$ is nonnegative and presented as a sum of squares; and our results imply the lower bound $\Omega(\log(1/\varepsilon)/\gamma\sqrt{\delta})$ for this case.
Read the full abstract ▾
Machine Learning
When Do Supervised UQ Ensembles Improve LLM Hallucination Detection? A Robustness Study
Mohit Singh Chauhan, Vipin Gyanchandani and Dylan Bouchard
Uncertainty quantification (UQ) methods are widely used for
hallucination detection in large
language models (LLMs) in closed-book settings where ground-truth evidence is unavailable at inference time. Prior work has proposed combining UQ signals via learned ensembles, but empirical investigations into the robustness of these ensembles are limited. We study a supervised ensembling framework that trains a classifier over heterogeneous UQ-based scorer outputs on a small, domain-specific
dataset of labeled LLM responses, then applies it to out-of-sample
hallucination classification without
retrieval, tools, or reference documents. Across four LLMs, nine
datasets, and three generation regimes (short-form QA, long-form generation, and code generation), we provide a systematic robustness analysis along three axes: sample efficiency, in-domain
dataset transfer, and generation regime dependence. We find that supervised ensembles outperform the best individual scorer in 30 of 32 settings, with gains realized from as few as 100 labeled instances. Ensembles retain most of their advantage in cases of in-domain transfer under distribution shift, outperforming the best non-ensemble scorer in 23 of 28 transfer settings. Sampling-based black-box ensembles are nearly as effective as full ensembles, while single-generation white-box ensembles offer limited benefit.
Read the full abstract ▾
Computational Engineering
Concept of Time-Reversal Characteristic Modes in Non-Free-Space Environments
Chenbo Shi and Jin Pan
Characteristic modes possess a natural environmental interpretation in the current domain because the surrounding scene is carried by the Green function used to construct the impedance operator. An equally direct physical interpretation is less evident in the scattering domain once the background itself participates in propagation and feedback. This paper introduces a field-level definition of time-reversal characteristic modes based on the differential scattered field relative to a prescribed background. A characteristic state is identified when this additional field, after time reversal and propagation through the same background, regenerates the same differential scattering state up to a scalar modal factor. The concept is verified in two distinct non-free-space settings. For a finite structural background, it reproduces established substructure characteristic modes and agrees with a common-origin spherical-wave realization. For a PEC half-space, it agrees with the modes obtained from the half-space Green-function impedance operator, although the infinite plane has no natural finite-object transition
matrix. The results show that the physical modal statement can remain unchanged even when the numerical representation of the environment is fundamentally different.
Read the full abstract ▾
Machine Learning
Where Entropy Is Measured Matters: Policy Geometry in Bounded Continuous-Control PPO
Yiyang He, Zhichun Zhou, Ziwei Wang et al. (+2 more)
Many continuous-control policies are optimized as unbounded Gaussians and then mapped into bounded actions. We show that where
entropy is measured changes the policy geometry learned by proximal policy optimization (PPO). In an 80-muscle MyoLeg task, a clipped Gaussian executes 89.07% of actions within 5% of a bound. A same-state decomposition shows that this is not due to variance alone: setting variance to zero still leaves 83.83% of actions near a bound, while 82.12% of state-conditioned means lie outside the executable interval. Replacing clipping with a tanh map does not remove the high-variance regime. For latent Gaussian
entropy H(u), the
entropy loss has zero gradient with respect to the mean and a constant variance-increasing gradient. For executed-action
entropy H(a), the transform Jacobian adds an inward gradient on the mean. Across three matched MyoLeg seeds, near-boundary occupancy is 71.42%, 29.76%, and 18.83% under latent
entropy, no
entropy, and executed-action
entropy. A 38-dimensional Dog-Stand replication with an independent CleanRL-based PPO implementation reproduces the ordering in mean geometry, which also survives shared-state evaluation and boundary margins from 1% to 10%. Direct mean penalties can match or exceed the centering produced by H(a), showing that interior means are not unique to executed
entropy. However, matched mean geometry can coexist with substantially different variance and return.
Entropy measurement space is therefore a coupled mean-variance design choice, and task return alone does not characterize bounded-policy geometry.
Read the full abstract ▾
All authors Yiyang He, Zhichun Zhou, Ziwei Wang, Tao Xue, Haolin Fei
Computer Vision · Cross-listed from eess.IV
Unified-protocol voxel-level pulmonary embolism annotations for three public CT angiography datasets
Qihang Sun, Zhongxiao Liu, Bailiang Jian et al. (+6 more)
Reliable clot-volume quantification and subsequent risk assessment in pulmonary embolism depend on precise
segmentation of emboli on computed
tomography pulmonary angiography. Deep learning models for this task must be trained on accurate voxel-level labels. The three public
datasets that provide such labels were annotated under different protocols, and some of their studies contain unlabeled emboli or labels that are discontinuous across slices. This Data Descriptor presents voxel-level pulmonary embolism annotations for 149 of the 166 studies in these
datasets. A primary rater drew all annotations under a single protocol. A thoracic radiologist with more than 20 years of experience reviewed and revised them. Three raters at three different centers independently annotated a subset of 15 studies. The subset was selected by source
dataset and embolus location. Technical validation quantifies volumetric agreement with the source annotations, changes in within-mask attenuation, and inter-rater agreement on the subset. The
dataset is intended to allow
segmentation models to be developed and compared under a common reference standard.
Read the full abstract ▾
All authors Qihang Sun, Zhongxiao Liu, Bailiang Jian, Shenman Qiu, Jingyuan Wang, Lei Zhang, Lixiang Xie, Jiazhen Pan, Christian Wachinger
Robotics
NeuralParker: A Reinforcement Learning Planner for Irregular Parking Environments
Zihan Wang, Bai Huang, Yang Guan et al. (+4 more)
Automated parking commonly assumes marked slots and short approach maneuvers. Delivery and service vehicles, however, may need to reach an operator-specified pose in an irregular bounded environment from a distant start. Existing learning-based parking planners often rely on local observations, which can restrict long-range route
reasoning. To address this problem, we present NeuralParker, a reinforcement learning-based hybrid planner for arbitrary-pose parking. NeuralParker encodes full-environment obstacle and boundary geometry in a target-relative vertex representation, allowing the policy to retain route-defining context throughout the approach. It further couples a learned
curvature–length arc policy with an in-loop terminal ensemble that selects from diverse cubic Hermite connections using a
curvature-regularized cost. We also establish factorial and long-range route-choice
benchmarks to evaluate planning success and trajectory quality. Experiments on these
benchmarks show that NeuralParker achieves higher planning success and better overall trajectory quality than the evaluated baselines, while ablation studies support the benefits of the target-relative global representation and terminal ensemble. Finally, a real-vehicle evaluation confirms that the planner transfers effectively to real delivery-vehicle perception at a working parking site, planning successfully at low computational cost.
Read the full abstract ▾
All authors Zihan Wang, Bai Huang, Yang Guan, Xiao Li, Haoyu Xu, Naizheng Wang, Shengbo Eben Li
Machine Learning
Beyond Static Interpretability: Anticipating Post-SFT Mechanisms from Pre-SFT Parameters for Better Tuning
Hang Chen, Jiaying Zhu and Wenya Wang
Mechanistic Localization bridges mechanistic interpretability and post-training optimization by isolating critical parameters via interpretative approaches and then guiding parameter-efficient Supervised Fine-Tuning (SFT) in a “locating-then-tuning” paradigm. However, due to the retrospective nature of mechanistic interpretability, directly interpreting pre-SFT models introduces misleading conclusions. Specifically for novel tasks, initially identified
neurons differ drastically from those governing the final model, introducing biases that actively disrupt SFT. To address this, we propose a forward-looking localization framework that accurately estimates the post-SFT interpretability state using only pre-SFT parameters and the target
dataset. Theoretically, we model SFT as a continuous parameter
evolution, leveraging Taylor expansion to rigorously bridge the post-tuning mechanistic objective with the pre-SFT model's dynamic gradients. Practically, we design dual-granularity (
neuron- and component-level) localization pipelines. Extensive experiments demonstrate that our approach not only provides superior SFT guidance but also exhibits robust performance and temporal scalability across increasing model sizes. This work transcends the fundamental limitation of traditional interpretability-its inability to identify task-critical mechanisms before they are trained-pioneering a predictive frontier that unites mechanistic interpretability with targeted optimization.
Read the full abstract ▾
Neural and Evolutionary Computing
On Scaling Coordinate-Based Neuroevolution: The Quadtree Bottleneck in ES-HyperNEAT
Romain Claret, Michael O'Neill, Paul Cotofrei et al. (+1 more)
ES-HyperNEAT evolves substrate
topology through adaptive quadtree subdivision; to our knowledge, no implementation with full population-level GPU parallelization exists. We present JAX-ESHN, a JAX-based implementation targeting GPU parallelization with batched CPPN queries, and
benchmark it against the CPU-based PUREPLES Baseline across five tasks: XOR, Parity-3, circle classification, sine regression, and CartPole. The core limitation is structural: each CPPN discovers a unique set of substrate positions, preventing population-level vectorization via vmap. On XOR, the CPU Baseline's runtime scales exponentially with depth while JAX-ESHN's construction cost on GPU (compilation plus first-generation evaluation) plateaus at deep substrates, so JAX-ESHN solves reliably where the Baseline rarely succeeds, with lower runtime variance. A CPU-vs-CPU multi-
benchmark control reproduces the same scaling divergence across Boolean, continuous, and control task types, confirming it is a property of the substrate-discovery implementation, not of GPU hardware. An alternative data structure (Hierarchical Spatial Hash Grid) fails not because it precomputes positions but because it applies the variance test independently per position, discarding the quadtree's parent-gated filtering and with it the adaptive sparsity essential to ES-HyperNEAT. These findings define the structural constraints any substrate-discovery method must satisfy to scale coordinate-based neuroevolution; the companion EMR-HyperNEAT reformulation, which replaces adaptive subdivision with eager evaluation of a static multi-resolution grid, satisfies them and resolves the bottleneck this paper characterizes.
Read the full abstract ▾
All authors Romain Claret, Michael O'Neill, Paul Cotofrei, Kilian Stoffel
Machine Learning
WarpSAC: Towards the Pinnacle of Scalable Off-policy RL by Rethinking Exploration and Exploitation
Zihao Wu, Hongyao Tang, Yi Ma et al. (+9 more)
Massively parallel simulation changes the data regime in which off-policy reinforcement learning (RL) is trained, challenging stabilizers designed for data-limited replay. Through controlled experiments across eight
benchmark families, we show that these stabilizers are data-regime-dependent: parameter normalization helps with narrow replay coverage but restricts value fitting when data are abundant, while clipped double-Q can be relaxed in high-throughput manipulation. Age-biased replay weighting improves learning efficiency across regimes, especially with limited network capacity. Based on these findings, we propose WarpSAC, a regime-aware family of off-policy RL
algorithms. WarpSAC uses Sample Weight Decay for efficient exploitation and provides two variants: WarpSAC-L (Norm ON, clipped double-Q) for data-limited CPU-scale training, and WarpSAC-A (Norm OFF, single-Q) for data-abundant GPU-parallel training. WarpSAC improves normalized score–step AUC over FlashSAC by 4.5% across nine CPU-scale environments and 23.1% across fourteen GPU-parallel environments. It increases UnitreeG1TransportBox-v1 success rate from 19.8% to 96.4%, improves mean normalized wall-time AUC on MuJoCo Playground by 19.1%, and achieves 36.4% faster sim-to-real deployment on Unitree G1 than FlashSAC. These results show that scalable off-policy RL should adapt its stabilizers to the available data regime.
Read the full abstract ▾
All authors Zihao Wu, Hongyao Tang, Yi Ma, Huizhong Song, Pengyi Li, Yifu Yuan, Fei Ni, Jinyi Liu, Wei Wei, Jianrong Wang, Yan Zheng, Jianye Hao
Programming Languages
NoC-Out: A Formally-verified Network-on-Chip Library for Rule-based Hardware Designs
Max Kurze, František Farka and Sebastian Ertel
The Network-on-Chip (NoC) is the communication backbone of any multiprocessor chip. A failure of the NoC has severe consequences for the whole system. Yet, no approach exists that provides formally-verified NoCs with strong guarantees but without tedious verification effort. Any library that generates formally-verified NoCs needs to be parametric in the structure of the NoC. This requires a hardware description
language (HDL) that allows for parametric and concurrent yet efficient hardware designs as well as the necessary program logic to reason about them in a modular fashion. So far, HDLs fall short in both aspects. In this paper, we implement NoC-Out, the first library/generator for formally-verified k-dimensional NoC designs. In order to build NoC-Out, we extended Kôika, a rule-based HDL in the Rocq
theorem prover, with support for concurrent yet efficient NoC designs and a program logic for modular, automated
reasoning. Given a configuration, NoC-Out produces a k-dimensional
torus NoC in Kôika, which can then be compiled to Verilog. Each produced NoC is equipped with a proof that it refines our formal NoC specification; no additional verification effort is required. Our specification proves a strong liveness guarantee, which consequently applies to all generated NoCs. In our evaluation, we find that our verification approach is even required to synthesize efficient NoCs in rule-based HDLs.
Read the full abstract ▾
Computation and Language
Ana Gjorgjevikj, Barbara Koroušić Seljak and Tome Eftimov
Multilingual text
embedding models enable cross-lingual transfer of knowledge across a wide range of NLP tasks, but their evaluation remains highly uneven across high-, mid- and low-resource
languages. In this paper, we propose a two-dimensional framework, specifically tailored for analyzing
multilingual embedding benchmarks under
dataset scarcity, and apply it on the Slavic-
language subset of the MTEB
benchmark. The framework distinguishes between task-specific and cross-task evaluation, while jointly analyzing three complementary aspects: (1) ranking robustness, (2) model consistency, and (3) evidence strength. At the task-specific level, we evaluate the stability of model rankings under changes in ranking methodology and
benchmark dataset composition. At the cross-task level, we assess the ability of models to generalize across diverse tasks within a
language. To quantify the reliability of
benchmark conclusions, we introduce an Evidence Strength Score that accounts for
dataset availability, diversity, and robustness assessability. Our analysis reveals severe
benchmark sparsity, with many Slavic
language-task pairs relying on a single
dataset or highly correlated
benchmark collections, limiting the ability to draw robust conclusions. The cross-task analysis reveals a small group of highly transferable models, most notably llama-embed-nemotron-8b,
multilingual-e5-large-instruct, and Qwen3-
Embedding variants, that consistently perform well across Slavic
languages and tasks. Overall, the results demonstrate that
benchmark rankings and robustness conclusions must be interpreted jointly with certain notation of their evidence strength and highlight
benchmark scarcity as a major obstacle to trustworthy
multilingual evaluation.
Read the full abstract ▾
Software Engineering
Ockhamareto: Pareto-Gated Segment-Level Credit Assignment for Concise Unit-Test Generation with Reinforcement Learning
Dong Huang, Mark Harman, Jie M. Zhang et al. (+3 more)
We introduce Ockhamareto, a single-shot GRPO framework for unit-test generation and selection, based on the principles of Ockham's Razor and Pareto Optimality. Ockhamareto has two principal components: (i)~a Pareto-gated Bonus that rewards only rollouts non-dominated in~(mutation, $-$\#tests) space, and (ii)~
Token-level Segment Credit, which attributes each test's marginal mutation kills back to the
tokens of its unit-test block. On the UnLeakedTestBench~(ULT), Ockhamareto strictly Pareto-dominates the strongest RL baseline~(MIST-RL). Furthermore, it dominates on each and all optimization objectives, catching more bugs ($49.9\%$ vs $31.3\%$ mutation score at $N{=}5$), using fewer tests ($2.60$ vs $4.67$ on average), thereby achieving $3.4\times$ the per-test trade-off improvement. The advantage is found in all four
benchmarks~(HumanEval+, MBPP+, CodeContests, TestGenEval-Lite): Ockhamareto leads both mutation and coverage metrics on every one, always with the smallest suite. Ockhamareto also outperforms the state-of-the-art at all model scales, adding $+30$–$35$~pp mutation at 4B, 9B, and 27B model sizes. We also show that the knee point of the optimal trade-off between efficiency and effectiveness on the Pareto front is not correlated with obvious more easily computed proxy metrics, such as function size. This finding motivates the Pareto front computation; it is needed to identify this crucial engineering trade-off for each function under test.
Read the full abstract ▾
All authors Dong Huang, Mark Harman, Jie M. Zhang, Zhijiang Guo, Mingzhe Du, See Kiong Ng