Papers

Recent work on multi-agent LLM systems reaches sharply different conclusions: some results show that a single agent with the same information and compute should dominate a delegated system, others that multi-agent gains grow with task depth. We argue that much of the disagreement comes from modelling different bottlenecks, and introduce a stylised reliability model built around two trade-offs. Decomposition reduces the burden of long contexts but incurs a handoff tax when information is compressed or transferred between agents. Redundancy gains from multiple samples, but its benefit depends on how much their failures are shared. With reasoning budget, verification, and task structure added, the model yields two crossover conditions: decomposition becomes preferable once the attention cost avoided by resetting context exceeds the handoff cost, and parallel sampling at equal budget is eventually preferable when its shared-failure floor lies below the error floor of one agent thinking longer. We connect these regimes to recent theoretical and empirical results. On a ledger-reconciliation task we measure the context-degradation curve and the handoff tax from single-agent and handoff runs alone. From these the model places the crossover at depth 10 and predicts decomposition to win at depths 20, 50, and 100. It does, on step-level and final-balance accuracy, and the decomposed system's success, which the prediction never sees, lands within 9 percentage points of the predicted rate at every depth.

Boscaggin, Feltrin and Zanolin conjectured that the Neumann and periodic problems associated with the equation \begin{equation*} u”+a(t)u^\gamma=0 \end{equation*} with a two-step indefinite weight $a(t)$ have at most one positive solution for every $\gamma\in\mathbb{R}\setminus\left\{-1,0,1\right\}$. We prove this conjecture and complete the uniqueness picture by also including the logarithmic potential case $\gamma=-1$. More generally, for every $p>1$, we consider the equation \begin{equation*} \left(\left\lvert u'\right\rvert^{p-2}u'\right)'+a(t)u^\gamma=0 \end{equation*} under Neumann or periodic boundary conditions and establish a complete existence and uniqueness classification for all $\gamma\in\mathbb{R}\setminus\left\{0,p-1\right\}$. The necessary mean-value condition $\gamma\int_{0}^{T}a(t)\,\mathrm{d} t<0$ is also sufficient for existence when $\gamma\leq-1$ or $\gamma>p-1$, while an additional sharp condition on the mean of the weight is required in the intermediate range $\gamma\in(-1,p-1)\setminus\{0\}$. Finally, a numerical example provides evidence of multiple positive solutions for the Neumann problem whose weight has the same single change of sign but a non-constant internal profile, highlighting the role of the two-step assumption.

We construct a complex structure on $T^2\times S^4$, giving an affirmative answer to the genus-one case of Calabi's problem. The resulting compact complex threefold is non-Kähler and admits no biholomorphic product decomposition.

Regional weighted phase-volume contraction for nonlinear systems is studied using positive densities selected from a finite-dimensional log-affine family. The affine dependence of the density divergence on the design parameters turns density selection into a semi-infinite convex optimization problem that maximizes a certified uniform contraction margin on a compact region. This provides a constructive alternative to testing a preselected density and ensures a well-posed design problem under compact parameter constraints. Finite active-point and minimax characterizations identify the worst-case states governing the optimum, while rigorous sampling bounds allow continuum contraction guarantees to be recovered from finitely many inequalities. For scalar power densities, feasibility is characterized by an exact interval condition. The same structure is preserved in discrete time and admits distance-weighted and conformal-geometric interpretations. Overall, the framework provides computable certificates for detecting regional contraction that may remain hidden from ordinary divergence. A two-parameter limit-cycle example demonstrates this effect when both ordinary phase volume and a natural single-basis density fail.

Living neuronal networks transform inputs through recurrent cellular and population dynamics, yet it is unknown which network architecture supports which computation. Neurons-on-a-chip turn this question into a design problem because microchannels guide axonal growth and set the network architecture. We introduce IC$^3$, an Integrated Characterisation of Communication-Driven Computation, which characterizes network state through neuronal dynamics, functional communication, and structural support. We implemented nine architectures in silico as conductance-based spiking networks and tested each on frequency decoding, temporal-order discrimination, and fading memory. Predominantly feedforward circuits decoded best. Sequential Chain and Microchannel Diode had the lowest IC$^3$ and recruited a third of reachable neurons, yet achieved the two highest scores on both classification tasks. Across architectures, higher IC$^3$ went with lower classification scores. We term this new direction neurotopomorphic computing, in which the physical organisation of neuronal connectivity is engineered as part of the computing substrate.

Large Language Model (LLM) assistants routinely decide whether they can trust users and third parties whose competence, intentions, and integrity they cannot verify. This uncertainty matters for safety, as trusting the wrong party can lead an agent to comply with harmful requests or act on malicious instructions encountered during tool use. To study this problem, we define trust as an assistant's willingness to accept vulnerability to the actions of another party and ask whether such behavior can be causally controlled through model activations. We build 2,000 contrastive conversations spanning ability, benevolence, and integrity, where paired responses complete the same request but differ in whether the assistant trusts the user. From these pairs, we learn steering matrices while keeping the model parameters frozen and test them across six instruction-tuned models from three families, finding that steering changes trust decisions monotonically in both directions. We then ask whether this effect extends to several safety-related agent settings involving harmful requests, prompt injections, and insider threats, while using benign-task and reasoning as controls. Our findings provide evidence that trust in the user can be causally controlled along linear directions in model activations and provide a way to study how trust shapes safety-relevant behavior in language models.

Semantic segmentation of street panoramas can support detailed descriptions of urban environments, yet small datasets and unequal training costs make model selection difficult. This paper documents the two stages of a submission to the PalmCity challenge. In the first stage, nine pretrained segmentation systems are compared using approximately equal computation budgets, and the two leading candidates are trained with three random seeds. An equal probability average of three Encoder only Mask Transformer models with DINOv3 features, evaluated at three image scales with horizontal reflection, reaches 60.95% mean intersection over union on the 84 image validation split and 57.08% on the hidden test leaderboard, first in the snapshot of 5 October 2026. In the second stage, color and crop augmentation, rare class sampling, larger inputs, further architectures, and post processing are tested under the same update budget, and candidates are compared through a paired image bootstrap with selection rules fixed before the results. Training at 768 by 1536 and 896 by 1792 pixels gives the largest gains. The selected ensemble of four models scores 64.53% on validation when trained on the training images alone. Retrained on the training and validation images, it receives 61.54% mean intersection over union and 71.93% mean F1 on the hidden test leaderboard, first among five entries in the snapshot of 30 October 2026. Segment Anything voting, instance mask overlays, and ensembles with other architectures all reduce the validation score. Because the validation images guided many choices and the final models were trained on them, the validation scores are optimistic and the test score is the only independent measurement. The study reports all runs, comparisons, costs, and limits, and releases the code and models.

Federated multi-player multi-armed bandit problems model collaborative sequential decision-making where multiple players interact with a common bandit environment and share information through a central server to accelerate learning. Existing federated bandit frameworks typically assume that all players willingly share their local observations with the server. However, this assumption is often unrealistic in practical settings where players are self-interested and may not participate in collaboration without explicit incentives. To address this challenge, we propose an incentive-aware federated bandit framework in which players receive rewards for sharing information with the server and incur costs when buying information from the server. We develop a UCB-based algorithm, termed Buying-UCB, that balances individual exploration and collaborative learning by incorporating both sharing incentives and information acquisition costs into the learning process. We theoretically analyze the proposed algorithm and derive upper bounds on the group regret and buying cost. Our analysis further characterizes the trade-off between fully collaborative federated learning and completely independent learning. Extensive numerical experiments validate the theoretical findings and demonstrate the effectiveness of the proposed framework under different collaboration and pricing regimes.

Trusted Execution Environments (TEEs) are secure foundations for protecting sensitive information and executing computations over confidential data. Processing in an isolated execution environment (enclave) is invoked by an untrusted application in the Rich Execution Environment (REE). Then the enclave returns the execution results to the untrusted application. Even if the enclave has been attested, the boundary between the enclave and the untrusted application poses risks, such as tampering with input arguments or return values and manipulating the order in which the application invokes functions. Therefore, it is necessary to establish the authenticity and integrity of not only the enclave itself but also the application and its execution environment. In this study, we propose an attestation protocol, Portable Remote Attestation TLS (PRA-TLS), for a client application that invokes an enclave. We introduce a daemon that acts as an attester. PRA-TLS uses a trusted Attester Daemon to measure the state of the host environment and application code at runtime, providing these measurements as attestation evidence for verification by a remote Verifier. This mechanism allows the enclave to proceed only when the software environment is in the expected state and the application code is verified as legitimate. We define attack models and security requirements for the proposed protocol and formally evaluate its security using the Tamarin Prover. Furthermore, we implement a prototype of PRA-TLS using Intel SGX and evaluate its performance.

User behavior simulation is widely used to evaluate interactive information retrieval systems, but classical state-based approaches (e.g., Markov models) have limited ability to incorporate contextual information relevant for decision-making. We address this limitation by introducing a feature-conditioned Markov-style user model, in which transition probabilities are modeled as functions of positional, content-based, and interaction-derived features, enabling context-aware decision making while preserving the structural simplicity and computational efficiency of state-based models. Applying a multi-level framework that assesses predictive fit and behavioral fidelity, we analyze how different sources of contextual information contribute to realistic user simulation across multiple datasets, search settings, and feature configurations. Our results show that incorporating contextual features improves the models' ability to reproduce key aspects of real user interactions, but that their effectiveness hinges on search scenario and modeling objective. Instead of a one-size-fits-all solution, effective simulation requires task- and setting-specific feature selection. Our framework provides a practical and interpretable basis for making these choices.

While independent and identically distributed noise is typically assumed in quantum computing research, complex error models such as correlated and spatio-temporal errors remain underexplored. In this paper, the performance of $\mathrm{XZZX}$ and rotated surface codes is evaluated under these complex noise models. Standard graph-based decoders struggle with the hyperedges generated by correlated errors on bias-tailored codes. To address this, we employ the Minimum Weight Parity Factor (MWPF) decoder, which natively processes syndrome hypergraphs without artificial edge decomposition. Our simulations show that under strong Z-biased circuit-level correlated noise ($\eta=100$), the threshold of the rotated code degrades to $\approx 0.36\%$, while the $\mathrm{XZZX}$ code drops to $\approx 0.70\%$. Furthermore, a distinct non-monotonic threshold dependence is observed for the $\mathrm{XZZX}$ code across different bias ratios. While the $\mathrm{XZZX}$ code effectively absorbs high-density Z-type hyperedges at extreme biases, it exhibits a threshold minimum at moderate biases ($\eta \approx 10$) because transverse X and Y errors actively bridge the diagonal strings, breaking the code's one-dimensional decoupling symmetry. These findings demonstrate that utilizing hypergraph-aware decoders is critical, as the performance of bias-tailored codes is highly sensitive to the geometric structure of spatio-temporal correlated noise.

RISC-V-based hardware/software co-design has become an established methodology for implementing post-quantum cryptography (PQC) on embedded devices. Its software programmability, extensible instruction-set architecture, and optional dedicated accelerators create a broad design space that offers the flexibility required for crypto-agility and post-deployment updates. For HQC, a PQC key-encapsulation mechanism selected for standardization, several RISC-V-based implementations have already been published. However, these works explore only a small subset of this design space. We present, to the best of our knowledge, the most comprehensive exploration of RISC-V-based HQC. Our exploration is centered around a new, parametrizable architecture framework on which we characterize 29,920 distinct architectural configurations across all three HQC security levels on FPGA. The Pareto-optimal configurations from this framework outperform the state-of-the-art RISC-V-based HQC implementation, ranging from 4.8x lower LUT utilization at 17% lower latency to 69.6x lower latency at 30% lower LUT utilization, while also requiring up to 10.6x fewer block RAMs.

Modern text-to-speech (TTS) systems can clone a target speaker from a short reference clip or be fine-tuned on a target voice, yet their behaviour on morphologically rich, lower-resource languages such as Turkish remain under-characterised. We present a systematic benchmark of four contemporary systems (Chatterbox, CosyVoice, OmniVoice, and VoxCPM2) evaluated across fine-tuned and zero-shot configurations, contrasted with a conventional VITS baseline and anchored to natural gold speech. Each configuration is scored with eighteen complementary objective metrics spanning learned naturalness predictors (UTMOS v2, DNSMOS-Pro, SCOREQ, WhisQA, AudioBox-PQ, NatScore, SpeechLMScore), intelligibility and signal-quality estimators (SQUIM PESQ/SI-SDR/STOI, Brouhaha), speaker similarity, distributional fidelity (TTSDS) and low-level acoustic descriptors. We further analyse how quality varies with utterance length and quantify long-form temporal consistency through speaker-identity and naturalness drift over chunked utterances. We release our evaluation code to support reproducible TTS evaluation.

Large Vision-Language Models (LVLMs) frequently suffer from object hallucination. Existing training-free interventions primarily manipulate attention weights, which indirectly affect the deep semantics reaching the final predictive layers. In this work, we shift our focus to the hidden state vectors extracted after self-attention and residual addition. Empirical analysis reveals that hallucinated tokens do not simply over-rely on linguistic priors; instead, they exhibit an anomalous contextual deviation, showing significantly lower similarities to both textual and visual contexts in intermediate layers. Motivated by this, we propose ROT, a layer-specific, training-free framework. ROT dynamically detects semantic deviation in the middle layers and applies a norm-preserving rotation to steer the hidden states back toward the local multimodal context plane spanned by the contexts. For subsequent layers, a representational smoothing mechanism is introduced to stabilize the calibrated trajectory. Extensive experiments on multiple benchmarks demonstrate that ROT consistently reduces hallucinations across various model architectures and scales, offering an efficient, geometry-driven solution for grounded generation.

Learning-based mutant selection is used to reduce the cost of mutation testing by ranking and selecting the mutants that seem to be the most promising. Prior work has shown great potential for this approach in fault revelation and subsuming-mutant selection. However, the risks and benefits of these approaches – especially the distribution of selected mutants across code locations and how that distribution affects behavioral diversity and fault revelation – remain unclear. We investigate this distribution and its effects, and introduce LinePool, a simple, model-agnostic stratification step that distributes selections across source lines while retaining the underlying ranking signal. Across two datasets and budgets (2% and 5% of the mutants), we find that learning-based selection concentrates the budget on a few code locations, and may thereby systematically introduce 'blind spots', untested code areas, allowing faults to escape detection. On larger programs, the selected mutants also exhibit substantial overlap in kill behavior. LinePool substantially increases spatial coverage and reduces kill-set overlap. It also improves fault revelation at tight budgets and makes selection more robust to imperfect or shifted ranking signals. Comparisons with clustering and established diversification methods show that none consistently outperforms LinePool across datasets, supporting its use as a simple diversification step. In general, our results suggest that behavioral diversity should be considered alongside effectiveness when designing and evaluating mutant-selection methods, and that exploiting structural elements of the program is one simple way to obtain such diversity while maintaining effectiveness.

This paper presents a deep learning-based full event interpretation prototype developed for the electron-positron Future Circular Collider (FCC-$ee$) using simulated $e^{+}e^{-}\rightarrow Z\rightarrow b\bar{b}$ events at $\sqrt{s} = 91 \ \mathrm{GeV}$. The study was carried out using the Innovative Detector for Electron positron Accelerators (IDEA) detector concept as case study for future application. The proposed graph neural network framework aims to simultaneously identify and reconstruct the hierarchical decay trees of $b$-hadrons in each event using information from reconstructed charged particles. Events are represented as graphs, where reconstructed charged particles form the nodes and pairwise observables are encoded as edge features describing the relations between particles. On a test sample, 90% of reconstructed decay chains contain exactly the correct set of particles, irrespective of their inferred decay hierarchy. Across a range of benchmark $b$-hadron decays, this clustering efficiency exceeds 93% for decay chains spanning up to three generations of decays.

In this work we consider the Maximal Independent Set (MIS) problem and the metric Steiner Forest problem in the sublinear time setting, under the adjacency/distance matrix query model. First, we give an algorithm that estimates the size of an MIS up to a multiplicative factor of $(1+\eps)$ using $\tO(n^{4/3}/\eps^2)$ queries. This improves the best previous algorithm by Mahabadi, Roghani, Tarnawski, and Vakilian (SODA 2026), which had a query complexity of $\tO(n^{3/2}/\eps^2)$. Via a reduction from that work, this would automatically imply the same improvement for the problem of estimating the metric Steiner Forest cost up to an $O(\log n)$ factor. However, as our second contribution, we consider the Steiner Forest problem directly and provide an algorithm with $\tO(n)$ query complexity that is very simple and does not proceed via MIS.

Three-dimensional (3D) left ventricular (LV) reconstruction from sparse cardiac magnetic resonance (CMR) imaging remains challenging due to inter-slice misalignment and insufficient local spatial information between slices. Global aggregation of contour features may obscure local contour-to-surface relationships. We propose Local2Mesh, a spatially localized contour-to-mesh framework that deforms a template mesh to reconstruct 3D LV geometry from sparse 2D contours without 3D mesh annotations. The framework introduces geometry-aware alignment to correct inter-slice misalignment and a plane-aware Local Router that routes contour features to template vertices using vertex-to-plane distances. Local and global contour features then jointly guide graph-based template deformation for 3D LV reconstruction. Experiments on two public datasets, M&Ms-2 and ACDC, demonstrate superior geometric reconstruction and functional estimation over existing methods. Zero-shot transfer from M&Ms-2 to ACDC demonstrates strong cross-dataset generalization. Reconstructed meshes also improve disease classification over sparse contours, supporting their utility for downstream cardiac analysis. These results demonstrate that combining geometry-aware alignment with local contour-to-vertex modeling improves LV reconstruction from sparse 2D contours and supports downstream cardiac analysis. The code is available at https://github.com/hwu918945-alt/loca2mesh.

Urban environments are increasingly exposed to complex, dynamic, and interdependent risks, ranging from traffic incidents and infrastructure failures to large-scale crises such as extreme weather events and emergency response scenarios. In such conditions, decision-makers and operators are required to act under time-critical constraints while relying on fragmented, heterogeneous, and often incomplete information. The lack of unified situational awareness, combined with limited visibility and disconnected systems, significantly affects response time, coordination efficiency, and overall operational effectiveness. In parallel, modern cities have deployed extensive sensing infrastructures, including surveillance camera networks, IoT devices, connected vehicles, and satellite-based observation systems. Although these technologies generate vast amounts of data, their exploitation remains limited due to the absence of integrated platforms capable of real time data fusion, intelligent interpretation, and intuitive visualization. As a result, a substantial gap persists between data availability and actionable intelligence, particularly in safety-critical and crisis management applications. To address this challenge, this paper presents an AI-driven XR situational awareness and operational platform designed for urban crisis management and smart mobility operations. The proposed system integrates data from city infrastructure, connected vehicles, and VRUs, enabling a comprehensive and real-time understanding of the urban environment. Through a unified operational dashboard and AR interfaces, the platform supports both centralized monitoring and field-level interaction. By enhancing perception, enabling cooperative awareness, and delivering context-aware information, the proposed approach improves decision-making, coordination, and safety across diverse urban scenarios.

Large language model (LLM)-enhanced recommender systems leverage rich item semantics to support personalized recommendation. However, semantic representations alone do not determine which historical behaviors reflect persistent preferences and which mainly indicate recent interests, leaving an important aspect of user understanding unresolved. Recent advances in LLM inference show that newly available information can be used to refine the internal state during inference, thereby improving subsequent predictions. Inspired by this principle, we propose MATE (Memory Adaptation with Temporal Evidence), an adaptive user modeling framework for LLM-enhanced sequential recommendation. MATE first evaluates each newly observed interaction from two temporal perspectives: whether it is repeatedly supported by historical behaviors and whether it is consistent with recent interactions. The resulting temporal evidence controls the updates of two user-specific memories, where the long-term memory conservatively preserves persistent preferences while the short-term memory rapidly adapts to recent interests. For each recommendation, a recent-context representation dynamically determines how strongly the two memories contribute to the current user representation. During offline training, next-item prediction is jointly optimized with temporal supervision, while during online adaptation, the shared model remains fixed and only the two user memories are updated from newly observed interactions. Experiments on MovieLens-10M, Amazon Luxury Beauty, and KuaiRec show that MATE improves mean NDCG@10 over the strongest external baseline by 7.0–13.2%. Further analyses support its ability to adapt to recent interests while retaining useful information about recurring earlier preferences.

Sparse autoencoders (SAEs) are increasingly used not only to interpret language models but also to intervene on their internal representations. We show that this creates a supply-chain attack surface: a maliciously modified SAE can induce attacker-chosen behavior when inserted into the forward pass of an otherwise unchanged language model. We introduce a decoder-only SAE backdoor that leaves both the underlying LLM and the SAE encoder frozen, restricting the attack to a single auxiliary component at a single insertion layer. Using code generation as a case study, we demonstrate high rates of unsolicited code insertion across three language models and a wide range of insertion layers, as well as trigger-dependent behavior conditioned on a prompt cue. We further evaluate the modified SAEs using HumanEval and selected SAEBench metrics. While attack effectiveness varies across models and layers, strong backdoor behavior can coexist with relatively small changes in several conventional SAE quality measures. These results establish that SAEs can carry behavioral backdoors without modifying the language model itself and should therefore be treated as security-sensitive components.

The variational derivation of two-dimensional von Kármán shell energies from three-dimensional nonlinear elasticity is investigated in the presence of constant pressure loads. On the one hand, the paper extends the existing analysis of pressure-loaded plates to the curved setting of shells. On the other hand, it complements the classical variational derivation of shell theories by incorporating a nonlinear pressure force.

Agentic AI assistants increasingly act on developers' behalf by modifying files, executing commands, and accessing external resources. These actions often require permission, yet little is known about how practitioners make permission decisions while still benefiting from agent autonomy. To address this gap, we conducted a sequential mixed methods study, interviewing 18 practitioners who use AI agents and then surveying 115 practitioners based on the interview findings. We find that practitioners often understand agent behaviour through what they can directly observe and review, while decisions, data use, and other activity behind the scenes remain less clear. This uncertainty also shapes permission decisions, which depend on the scope and risk of an action, whether it fits the task, familiarity with the agent, and the environment in which it operates. Practitioners respond by adjusting how closely they oversee agents, from setting limits in advance to monitoring execution and reviewing work afterwards. How much scrutiny they apply depends on factors such as trust, task importance, time pressure, and the consequences of an action. Our findings suggest that permission systems should make consequential actions easier to review, distinguish what an agent is allowed to do from what the user intended, make reversibility clearer, avoid treating repeated approvals as stable preferences, and distinguish rejecting a single action from rejecting an entire approach.

Magnetic navigation provides contact-free and line-of-sight-independent feedback for robotic systems, yet existing approaches typically rely on explicit pose estimation or direct use of raw magnetic measurements, making accurate control susceptible to modeling errors, measurement noise, and disturbances. This work presents MagServo, a hierarchical learning-based framework for robust 6-DoF magnetic servoing directly using the learned latent magnetic feature. MagServo learns uncertainty-resilient magnetic representations through masked reconstruction and captures state-dependent interaction dynamics between robot motion and latent magnetic transitions without analytical magnetic models or explicit Jacobian supervision. Based on the learned dynamics, a hierarchical controller combines nonlinear model predictive control for coarse approach with local Jacobian inversion for precise fine regulation. Extensive physical experiments demonstrate submillimeter and subdegree accuracy, achieving mean terminal errors of 0.386 mm and 0.479 degree for 6-DoF pose reaching. MagServo further outperforms a localization-based control baseline in complex trajectory tracking and maintains robust performance under unseen magnetic-source configurations without retraining. A supplementary video of the real-robot experiments is available at https://youtu.be/rZt1NUP1Mr0.

The evolution of weighted phase volume under a fixed dynamical system is investigated using a positive weight raised to an arbitrary real exponent. Sufficient conditions for uniform exponential contraction and expansion of transported weighted volume are obtained in terms of the corresponding weighted divergence. These conditions make it possible to reveal dissipative properties that may remain undetected by the ordinary divergence. Consequences for invariant sets are established, and, under the assumption of a compact absorbing set, the weighted measure of a global attractor is characterized. The approach is extended to invariant submanifolds through a distance-weighted ambient-volume estimate. A differential-geometric interpretation is also provided, in which the weighted divergence is identified with the ordinary divergence associated with a conformally transformed metric. The dependence on the exponent is used to reveal a duality between mutually inverse weights. In addition, sufficient local conditions for exponential attraction to an invariant hypersurface are derived using a sign-changing weight. Weighted-volume decay is interpreted as an integral property of transported sets and, by itself, is not identified with pointwise Lyapunov stability.

Liquid-rocket thrust-chamber design involves interdependent analyses in which downstream constraints can require earlier design decisions to be revisited. Managing these dependencies across heterogeneous tools requires consistent design information and coordinated updates throughout the workflow. We present RocketAgent, a long-horizon engineering agent for multidisciplinary preliminary design of liquid-rocket thrust chambers. A single plan-owning Coding Agent coordinates engineering skills for performance sizing, subsystem optimization, geometry generation, and multiphysics assessment. A provenance-aware knowledge graph supports method selection, while a typed Design Intermediate Representation maintains shared parameters, artifacts, and decisions. Revision-aware checks invalidate affected results and block superseded inputs, with consequential changes subject to engineering approval. In a representative simulation-based design, RocketAgent continued from an infeasible cooling search through an engineer-authorized operating-point revision, identified feasible subsystem designs, and coordinated subsequent geometry generation and multiphysics assessment to support final configuration selection. Separate module tests assessed surrogate predictions and nozzle adaptation. A two-configuration comparison across three controlled scenarios verified the expected dependency invalidations and superseded-input blocking before solver execution. The representative case demonstrates sustained coordination across a multidisciplinary design workflow, while the controlled tests establish the behavior of the revision mechanisms supporting that execution.

In this paper, in the framework of the Biréfringence Magnétique du Vide (BMV) experiment, we present our current apparatus in detail and we discuss its critical points. We also report the results obtained using specially designed pulsed coils called Foil Coils. Our novel measurement of the vacuum linear magnetic birefringence constant is $K_{VMB} < 1.7 \cdot 10^{-21}~\text{T}^{-2}$ at a confidence level of 68$\%$, which is an improvement with respect to our previous 2014 measurement.

Discovery of causal relationships in high-dimensional Gene Regulatory Networks (GRN) is computationally challenging and often difficult to interpret due to dense connections. Therefore, grouping genes together into functional modules can improve tractability and biological interpretability. However, existing cluster level causal discovery methods assume access to a predefined admissible partitions, requiring the graph over clusters to be acyclic. Constructing such partitions is therefore challenging. In this work, we introduce GO-based Clustering for Causal Discovery (GO4CD), an algorithm that uses Gene Ontology (GO) to construct biologically meaningful gene partitions at multiple levels of granularity, while favoring those more likely to be admissible for causal discovery. GO4CD groups together genes participating in a shared biological process, and propagates gene annotations through the ontology hierarchy to achieve different granularity of partitions. Furthermore, we integrate GO4CD with Causal Learning over Clusters (CLOC) algorithm and evaluate recovery of true Markov equivalence class both with an oracle of conditional independencies and on simulated gene expression data using multivariate conditional independence tests. We evaluate GO4CD on multiple E.coli regulatory subnetworks and find that it is inadmissible in 18.1% of the cases, compared with 65.3-82.3% for the semantic-similarity baselines. Our results indicate that GO4CD is substantially better suited to learning causal GRNs defined over biologically meaningful gene clusters.

Deep learning has advanced chest X-ray (CXR) diagnosis, yet demographic biases in learned representations may contribute to performance disparities across intersectional groups. We propose a single-encoder framework combining dual-level decorrelation with prototype-guided cross-group contrastive learning to reduce demographic dependence while accounting for within-class variation. We further propose Demographic Representation Alignment Reduction (DRAR), a new metric that quantifies the reduction in demographic structure within disease representations. The framework is evaluated on four classification tasks using 34,809 CheXpert test images across eight intersectional groups, defined by age, sex and ethnicity. Compared with empirical risk minimization (ERM), our method reduces the mean equalized-odds gap from 15.41% to 10.86% and the AUC gap from 5.95% to 5.01%. Our method achieves a DRAR of 59.04% relative to ERM, with only a slight decrease in mean AUC. These results demonstrate that representation disentanglement can reduce demographic bias and improve intersectional fairness. Code is available at \url{https://github.com/06Yujie/Fair-Medical-Imaging}.

We compute the NNLO QCD contribution to the $N$-jettiness soft function for processes with two massive and an arbitrary number of massless final-state hard partons. The cancellation of all divergencies is demonstrated analytically. The finite remainder of the soft function is calculated numerically without an infrared regulator. The calculation employs the recently proposed method that uses the so-called inclusive $N$-jettiness variable to simplify the computation of the non-iterative contribution to the soft function.