Papers

Most video editing methods focus on changing the appearance of the source video, while offering limited control over its dynamics. We introduce Reimagine Video Dynamics (RVD), a framework that disentangles a compact, editable dynamics token from visual context. We learn this token through self-supervised reconstruction: given the first frame as visual context, a renderer must recover the original video from the dynamics token, encouraging it to capture how the scene evolves rather than how it looks. This disentanglement allows video dynamics to be edited directly while preserving visual context. We develop a language-guided dynamics-token editor that transforms source dynamics into target dynamics, and train it with a scalable counterfactual video-pair pipeline and a two-stage training strategy. Extensive experiments show that RVD enables effective video dynamics editing, training-free retiming, and appearance-controlled re-rendering.

Forward kinetostatics of spatial tendon-driven continuum robots typically requires a nonlinear equilibrium solve for each actuation input. This paper develops a force-to-Cartesian-configuration model with a closed-form solution in quadratures for spatial multi-segment robots under tendon actuation. The Cartesian backbone centerline and accumulated material twist serve as generalized coordinates, from which the strain measures and tendon geometry are derived. Variational equilibrium yields explicit axial and bending relations and establishes zero equilibrium material twist within the proposed model for admissible longitudinal non-helical routing. The solution is propagated segment by segment without an iterative equilibrium solve, while retaining axial deformation, spatially varying axial and bending stiffnesses and tendon-routing diameter, and segment-dependent tendon participation. Numerical comparisons with a full-strain geometric variable-strain model (GVS) yield maximum length-normalized tip-position discrepancies of 8.91 x 10^-6 and 1.01 x 10^-5 for the single- and three-segment robots, respectively. Mean evaluation times of 1.52 {\mu}s and 2.94 {\mu}s, with corresponding speedups of approximately 1864x and 3348x over the baseline, demonstrate the computational advantage of the explicit force-to-configuration mapping in the reported benchmark.

Provenance-based intrusion detection systems (PIDSs) identify suspicious activity in audit streams, but their outputs remain difficult to turn into coherent attack narratives. Direct LLM analyses of local anomalous subgraphs lack deployment-specific normal-behavior knowledge and validated attack state across evidence fragments. This can cause unsupported attack interpretations of routine activities and incorrect attribution of temporally dispersed evidence to attack stages. We present ANCHOR, an investigation-oriented provenance system that combines evidence curation with context-grounded LLM reasoning. It calibrates anomaly judgments by relation type and links anomalous windows through rare relation-role patterns. The resulting evidence queues preserve causal structure, temporal boundaries, and cross-window continuity. The investigator interprets process-centered evidence using two complementary forms of context. Deployment Context combines environment-specific interaction and object baselines with high-risk security knowledge. Case Context uses a confidence-gated Attack-Tracking Cache to maintain investigation state across windows. Correlating current evidence with high-confidence prior findings, ANCHOR incrementally reconstructs attack narratives organized by kill-chain stages. We evaluate ANCHOR on six DARPA Transparent Computing E3/E5 datasets across three operating systems. Controlled evidence-level and end-to-end comparisons show improved overall IoC recovery and attack-stage attribution over state-of-the-art provenance-based baselines. These gains persist under a fixed LLM backbone in our evaluation. ANCHOR processes a full audit day at dollar-level API cost, supporting practical, context-grounded investigation across windows.

Here we report the growth of single crystalline ScB2, an ultrahigh-temperature ceramic, at ambient pressure in a laser-heated Optical Floating Zone via the travelling solvent method. Crystals have been grown from both Sc-rich (55-65 at% Sc) and B-rich self-flux (80-83 at% B) at growth rates in the range of 0.2-2 mm/hr. The structure of grown crystals is in good agreement with an AlB2-type layered hexagonal phase, space group P6/mmm, with lattice constants a = 3.1423(2) Å\ (resp. 3.1502(3) Å) and c = 3.5084(3) Å\ (resp. 3.5041(3) Å) for crystals grown under Sc-rich (resp. B-rich) conditions. Crystals natively grow along the in-plane [100] direction. Electron backscattered diffraction shows that Sc-flux growth results in boules with multiple domains containing Sc inclusions, with the domains highly aligned. In contrast, B-flux boules are single domain after the initial nucleation region. Rocking-curve measurements of B-flux crystals for the (h000) and (000l) reflections show single peaks, establishing that the crystals are free from grain boundaries; the asymmetry in the scattered-intensity tails suggests the presence of point defects. Surface X-ray photoemission spectroscopy shows that the electronic environment in B-flux crystals is superior to that of Sc-flux crystals and produces highly resolved binding-energy peaks for B 1s and Sc 2p. Work-function measurements for the (11-20) plane give a value of approximately 5 eV, consistent with the highly electrically conductive nature of ScB2. These results demonstrate the viable ambient-pressure growth of ScB2, establish it as a lattice-matched substrate candidate for Al-rich AlGaN power microelectronics, and show that this growth route enables scalable manufacturing of ScB2 substrates.

We benchmarked vision-language models (VLMs) on the decisions annotators take when inspecting electron microscopy images in connectomics: synapse proofreading (presence and polarity) and neuron proofreading (split errors and merge errors). For synapse proofreading, we evaluated 19 open and 2 closed models across various architectures and sizes under zero-shot, four-shot in-context learning and LoRA settings, against specialist models, on datasets constructed by us using public resources. For neuron proofreading, we evaluated 3 open and 2 closed models on the ConnectomeBench2 dataset, with cross-species transfer from fly and mouse to human and zebrafish. Most models were at chance zero-shot; a few examples helped mainly the closed and largest open ones. LoRA on a few thousand labels brought open models level with specialist models. When evaluated on unseen species, the best adapted VLMs outperformed specialist models in identifying merge errors. The project will be publicly available upon acceptance.

This paper establishes the sharp lower bound $(4n\pi^2)^{n/2}$ for the Willmore energy $\mathcal{W}$ of flat $n$-tori in the Euclidean space. Up to Möbius transformations, the Clifford $n$-torus $\mathbb{S}^1\bigl(\sqrt{1/n}\,\bigr) \times \cdots \times \mathbb{S}^1\bigl(\sqrt{1/n}\,\bigr) \subset \mathbb{S}^{2n-1} \subset \mathbb{R}^{2n}$ is shown to be the unique minimizer attaining this bound. This also confirms Chen's conjecture for flat $n$-tori. However, when $n \geq3 $, we show that Chen's conjecture fails on the total mean curvature of general immersed $n$-tori: certain Möbius transformations of the Clifford $n$-torus strictly decrease the total mean curvature.

Latent space monitors aim to detect undesired behaviors in LLM agents by inspecting an agent's internal activations rather than its outputs. However, interactive monitoring creates a feedback channel where each verdict the monitor delivers leaks information to the model about how its internal states are being evaluated. We ask whether an agent can infer the monitor's decision rule from this feedback and then selectively edit its activations to evade detection. Unlike prior evasion attacks, the model is never explicitly told what the monitor detects. Surprisingly, off-the-shelf models already produce activation edits aligned with the monitored direction, but at insufficient magnitude for evasion. Simply scaling up these edits by a factor of 8 reduces the monitor's TPR from 100% to 27%. A rank-1 LoRA amplifies this behavior into effective evasion within the forward pass, reducing TPR further to 4% on held-out concept monitors while leaving other concepts at their normal detection rates. Capabilities on standard benchmarks are retained under this finetuning, and the evasion skill survives retraining the monitors on the new activations. Mechanistically, we find evidence that the model computes its activation edit from the prior in-context turns, and show that the edit becomes more aligned with the monitored direction as more examples are provided. These results demonstrate feedback-conditioned control over activations and suggest that latent monitoring should be treated as an interactive process in which agents can observe and respond to oversight measures.

It has remained a critical open question as to whether a longitudinal spin current gradient (e.g., due to a gradient in the thickness, composition, or width) can replace the longitudinal magnetic field required for deterministic spin-orbit torque switching of normal metal/ferromagnet heterostructures with perpendicular magnetic anisotropy. Here, we report robust micromagnetic and experimental evidence that any realistic longitudinal spin-current density gradient cannot be an effective replacement for a longitudinal magnetic field to enable deterministic switching of a perpendicular magnetic anisotropy device. Instead, the longitudinal spin current density can only modify the nucleation and pulse-timing-sensitive oscillations of magnetic domains and thus the picosecond-scale strip-like indeterministic switching windows. The same conclusions hold robustly when the transverse effective field associated with the spin-current pulse (the sum of the Oersted field and field-like torque) is taken into account. The indeterministic switching prevents applications in the presence of finite device-to-device variations and drifts in write current pulse and thermal fluctuations of device parameters. We also experimentally show that, instead of a longitudinal spin current density gradient, perpendicular spins are much more effective in deterministic switching of perpendicular spin torque devices.

Large language model (LLM) agents have demonstrated strong performance on complex web navigation tasks, yet they remain brittle in real-world settings where user intentions are underspecified and preferences are heterogeneous. In practice, users rarely provide explicit profiles, requiring agents to infer latent preferences from implicit signals. Despite its importance for deployment, this problem setting is largely underexplored in existing benchmarks. To address this gap, we introduce AdaptArena, a benchmark for evaluating test-time personalization of web agents via implicit preference inference. AdaptArena consists of 480 tasks, featuring both single-preference and double-preference scenarios. Each evaluation task must be solved by retrieving and leveraging the most relevant historical user trajectory that implicitly encodes the target preference. In addition, we introduce AdaptiveAgent, a retrieval-based framework for standardized evaluation of implicit preference inference. Experiments reveal a substantial performance gap: while oracle agents with access to ground-truth preferences achieve an 82.92% success rate, the evaluated LLM agents using our framework reach at most 15.62%. Furthermore, we find that correctly inferring user preferences is necessary but not sufficient for task success, as execution failures in downstream web interactions remain a significant bottleneck even when agents align with the target preference. These findings highlight implicit preference inference and robust action grounding as key challenges for deploying reliable, user-facing web agents. We release our code: https://github.com/McGill-NLP/web-agents-test-time-adaptations

Quantized tensor-train (QTT) compression is incorporated into a full-vector three-dimensional scattered-field finite-difference time-domain (FDTD) formulation on uniform Yee grids. All six electromagnetic-field components, material-dependent update coefficients, equivalent-current sources, and staggered finite-difference operators are represented in compatible QTT form. Gaussian regularization of voxelized material interfaces is used to reduce the coefficient ranks generated by abrupt dielectric and conductivity transitions. The formulation is evaluated for an anatomically heterogeneous human-head model and a homogeneous dielectric sphere on grids containing up to $512^3$ spatial cells. The reported results show that interface smoothing substantially reduces material-coefficient ranks and that the TT–FDTD solution reproduces the full-grid transient fields with pointwise absolute errors on the order of $10^{-4}$ in the examined slices. Compared with conventional FDTD, the tensor representation greatly reduces storage at fine discretizations, although tensor contractions and recompression introduce additional per-step computational cost. These results demonstrate the feasibility and memory–time tradeoff of QTT-accelerated three-dimensional FDTD for large structured-grid simulations.

We study an unknown-transition finite-horizon Markov decision process (MDP) with a finite collection of known reward functions $\{r^1, r^2, \ldots, r^M\}$. The goal is to output an $\epsilon$-optimal policy for every reward using online episodic interaction only. Performance is measured by the policy error $V_{0}^{*, m} - V_{0}^{\widehat\pi^{m}, m}$ where $m\in [M]$ represents the reward function and $V_{0}^{*, m}=\mathbb{E}_{s_1\sim \mu}[V_{1}^{*, m}(s_1)]$. Under this setting, we design a provably efficient algorithm to establish a minimax sample complexity bound of $$ O\left(\frac{SAH^3}{\epsilon^2}\log M \mathrm{polylog}\left(\frac{SAH\log M}{\min\left\{\epsilon, 1\right\}\delta}\right)\right)$$ episodes, with no additional burn-in cost. This matches the information-theoretic lower bound up to a factor of $ \mathrm{polylog}(SAH\log M/(\min\left\{\epsilon, 1\right\}\delta))$. Our method combines three technical ingredients. First, we adapt MVP to reward-switching learning to construct optimistic value estimates. Second, we use fresh replay samples to conservatively evaluate the candidate policies. Third, gap-based multiplicative weights updates adjust the reward-sampling distribution using the differences between these estimates, converting weighted learning progress into simultaneous guarantees for all rewards.

Modern Security Operations Centers struggle with delayed manual incident response, enabling adversaries to advance through the Cyber Kill Chain during early stage reconnaissance. While classical game theoretic defense models optimize strategic resource allocation, they rely on static utility matrices that fail to adapt to dynamic telemetry. This paper presents an integrated, metrics driven decision engine that bridges quantitative risk parameterization and continuous automated response time. Common Vulnerability Scoring Systems exploitability parameters are mapped to attacker success probabilities and evaluate defender log distributions via Factor Analysis of Information Risk Monte Carlo simulations. Real time SIEM logs streams are modeled as Poisson process arrival rates, dynamically updating defender posterior threat belief through sequential Bayesian filtering. A closed form threshold is derived by framing the interaction as a dynamic Bayesian Stackelberg game, where the expected unmitigated risk exceeds proactive containment cost. Parameterized against empirical data from the 2023 MGM Resorts and Caesars Entertainment cyber incident, simulation results demonstrate that the engine suppresses transient background noise while triggering automated SOAR network isolation within seconds of adversarial probing. Multi parameter sensitivity analysis confirms that the decision boundary dynamically adjusts to live perimeter vulnerability, offering a control theoretic foundation for sub minute automated threat containment.

On-policy distillation (OPD) trains a student to match the teacher's next-token distributions on the student's own trajectories and has yielded substantial empirical gains. Generalized variants allow the student to surpass the teacher by extrapolating an implicit reward in output space. The language-model head, however, attenuates this change anisotropically: much of the change encoded in the teacher's hidden states reaches the logits at a small fraction of its weight, and the sampled-token log-probability ratios on which output-space extrapolation relies inject noise that the extrapolation amplifies, making training unstable. We observe that reinforcement learning (RL) shifts a model's internal representations relative to its base checkpoint, and that the direction of this shift can be measured at every layer. Motivated by this observation, we propose RIDE (RL-Induced Direction Extrapolation), which extrapolates the RL-induced change directly in representation space: at every layer and token position, RIDE computes the residual between the teacher and its pre-RL checkpoint and regresses the student's hidden states toward targets displaced beyond the teacher along this residual. Conditioned on a sampled trajectory, this regression is equivalent to maximizing a linear directional reward defined by the residual under a quadratic penalty centered at the teacher, which makes explicit how the objective moves the student along the RL-induced direction while limiting its deviation from the teacher. Across four base/RL-teacher pairs spanning different scales, architectures, and pre-training lineages, RIDE approaches or exceeds the RL-trained teacher on every pair and is the only method whose mean does so, and it consistently outperforms output-space extrapolation, which degrades the student whenever the teacher is close to its base. Project page: https://github.com/xixixixixxxx/RIDE.

In this paper, we study a system of partial differential equations modeling the dynamics of nematic electrolytes on a two-dimensional torus $\mathbb{T}^2$. The model couples the Poisson–Nernst–Planck equations for the evolution of ion concentrations and electrostatic potential with the Ericksen-Leslie equations for the flow and orientation of nematic liquid crystals. We prove the global existence of weak solutions to this system. The proof relies on a Ginzburg-Landau approximation scheme. Key steps include deriving uniform energy estimates and establishing the strong convergence of the director field by utilizing a Pohozaev-type identity to handle the lack of compactness in the Ericksen stress tensor.

Quadrotor racing demands aggressive attitude and progress control while passing through every gate, and conventional quadrotor MPCC formulations state the prediction model in inertial coordinates and the attitude error in the body frame. We present a Dual-Quaternion Model Predictive Contouring Control (DQ-MPCC) for quadrotor racing in which the pose is a unit dual quaternion and the contouring errors are projected onto the tangent space of the dual quaternion manifold, expressed in the desired body frame: the same rigid-body dynamics as the conventional model, in unified pose-twist coordinates in the body frame. We compare DQ-MPCC against a baseline MPCC through Monte Carlo software-in-the-loop simulations and real-world racing on an eight-gate circuit of 11x4.5x3.65 m. With the same gains in simulation and hardware, DQ-MPCC keeps every crossing of its completed flights within the prescribed geometric margin, whereas the baseline exceeds it, its median worst-gate offset growing by 71.5% sim-to-real against a 10.1% decrease for DQ-MPCC. Among the configurations that keep every simulated gate crossing within the geometric margin, DQ-MPCC also reduces the minimum lap time by 6.7%, and by 10.5% in the real-world flights, while running onboard at 100 Hz.

Evidence shows that humans and AI systems perform better together, by collaborating, than alone. This paper examines two key design dimensions of human-AI collaboration (autonomy and initiative) and explores the collaboration patterns that they generate. Documenting these patterns starts with identifying the underlying problems and solutions, followed by examining the internal tensions within the problems. The paper uses a paradox perspective to analyze those tensions. It describes a process for surfacing the tensions and mapping the underlying paradoxes. It also illustrates how the pattern descriptions can be derived from mapping these paradoxes. Finally, the paper documents four human-AI collaboration patterns: Instruction, Delegation, Assistance, and Co-creation.

Decomposing finite Abelian black-box groups into cyclic factors is a basic problem in quantum computation. We give a quantum algorithm for decomposing finite Abelian black-box groups by adapting the quantum sampling and classical lattice-reduction method of Regev's factoring algorithm. The algorithm computes an invariant-factor decomposition and corresponding cyclic generators with high probability. For a group G of order at most 2^n with unique encodings and reversible group-operation cost T_op = Omega(sqrt(n)), our algorithm uses O(sqrt(n)) quantum circuits, each with O~(n T_op) gates and executed at most O(n) times. For comparison, we consider the Cheung-Mosca algorithm, which decomposes the Sylow p-subgroups separately. We also consider the common-modulus implementation of the extended Cheung-Mosca algorithm. Our algorithm reduces the sum of the circuit gate counts from O~(n^2 T_op) to O~(n^(3/2) T_op). The total quantum time bound reduces from O~(n^3 T_op) to O~(n^(5/2) T_op), and the quantum space bound reduces from O(n^2) to O(n) qubits. The classical computation uses polynomially many bit operations and group-operation queries. These improvements rely on two technical ingredients. We prove that the integer relation lattices used in our algorithm admit, with high probability, integral bases whose vectors have Euclidean norm at most exp(O(sqrt(n))). We preserve the circuit-size advantage of Regev's algorithm for group decomposition by incorporating all O(n) sampled generators while keeping the lattice-reduction dimension at O(sqrt(n)).

Quantum computing has already been explored in several network-security applications. However, how quantum computing may contribute to network-security classification in both the near term and the longer term has not been systematically discussed. This paper studies this question through two complementary experiments. First, we evaluate near-term quantum-kernel support vector machines (SVMs) on practical network-security classification tasks and compare them with classical SVM baselines on KDD Cup 1999, CICIDS2017, and BoT-IoT. Across these runs, quantum kernels are competitive. They can match or improve classical baselines in some settings, while classical RBF kernels remain stronger in others. This suggests that near-term quantum-kernel methods should be evaluated as practical, dataset-dependent alternatives to classical kernels rather than as uniformly superior replacements. Second, we use quantum oracle sketching (QOS) to study a longer-term memory advantage for classification with streaming classical samples. In QOS, samples are processed online and used to incrementally construct an approximate quantum oracle, which provides coherent query access for downstream quantum algorithms without retaining the entire dataset. Under the QOS-inspired machine-size estimate, comparable accuracy corresponds to a substantially smaller effective memory-size proxy than explicit sparse/QRAM-style storage. Compared with a simple streaming proxy, the result is more nuanced because aggressive feature filtering can make the streaming dimension small. This suggests that the long-term value of quantum computing for network-security classification may lie in memory-efficient data access rather than immediate runtime speedup. Together, these experiments show how quantum computing may contribute to network-security classification from near-term classification performance and longer-term memory efficiency.

The tragedy of the commons poses a multi-agent safety problem: reward-seeking agents can deplete a shared resource, and cooperation among its users does not itself specify how much must be preserved. We make preservation an explicit requirement by formulating a regenerative commons as a constrained Markov game or a constrained multi-agent MDP with a designer-specified depletion budget. We develop a nonstationary Lagrangian framework that constructs a policy sequence from solutions of unconstrained games or cooperative control problems. Extending earlier time-average constructions, we introduce average-epoch solution concepts for reset episodes with discounted rewards and terminal costs. We prove a reward-independent feasibility certificate, cooperative feasibility and approximate optimality against feasible policy mixtures, and an extension to unbiased sampled costs. For self-interested agents, a constrained Nash certificate quantifies the price-dispersion term introduced by deviations that redistribute budget across epochs. Under the stated assumptions on solver accuracy and multiplier updates, these results give constrained policy-sequence guarantees using solutions of unconstrained problems. Experiments with constrained IPPO and MAPPO in a Gordon-Schaefer fishery examine how depletion budgets shape stock retention, harvest rewards, and price adaptation.

Guard models are used as safety classifiers, with confidence scores driving downstream moderation decisions. We evaluate five guard models for prompt classification and find that although several are nearly calibrated on clean inputs, adversarial attacks degrade their calibration by an order of magnitude, turning false negatives into high-confidence errors indistinguishable from correct detections. Comparing each guard with its corresponding base LM, we find that uncertainty signals often remain available, with the base model typically expressing uncertainty on the same inputs where the guard fails. Layer-wise analyses localize this guard-base divergence to later layers, where guard models exhibit sharper safe/unsafe separation and lower-rank representations, while adversarial harmful inputs lie closer to the clean-safe region. These findings highlight a mismatch between guard confidence and base model uncertainty under attack.

We derive the finite-temperature characteristic function of work (CFW) for a driven $O(N)$ Gaussian field theory and obtain closed-form expressions for all zero-temperature excess-work cumulants. For gapped protocols, we use adiabatic perturbation theory (APT) to derive the $\tau_Q^{-2}$ scaling for protocols with a nonzero first derivative at either boundary, where $\tau_Q$ is the protocol duration; smoother boundaries lead to faster decay. For power-law protocols approaching the critical point with exponent $p$, we determine the competition between critical excitations and the regular contribution. The $n$th cumulant's critical contribution follows Kibble–Zurek (KZ) scaling $\tau_Q^{-p(d+n)/(p+2)}$ for $d+n<2p+4$, acquires a logarithmic correction $\tau_Q^{-2p}\log\tau_Q$ at $d+n=2p+4$, and scales as $\tau_Q^{-2p}$ above this condition, where $d$ is the spatial dimension. This analytical study of the APT–KZ crossover in a solvable model provides a basis for studying their competition in interacting field theories.

Cross-modal associations are systematic pairings of features across modalities, such as the association of 'bouba' with round shapes and 'kiki' with sharp shapes. Prior work has compared humans and vision-language models (VLMs) on such associations, but often using different stimuli or tasks between humans and models. Here, we ask whether VLMs align with humans not only in choices, but also in where they look when making those choices. We study both VLMs and humans (N = 53), presenting them with the same stimuli, a pseudo-word and two images, and record participants' choices and eye movements, which we release. We find choice alignment in a few larger VLMs, but their saliency matches human gaze less closely than a center-bias baseline, a fixed Gaussian at the center of each image. Fine-tuning small VLMs on human choices brings their choice alignment to the level of a human majority-vote reference on unseen words and images, yet their attention still matches human gaze less closely than this baseline. Training model attention on human gaze raises attention-gaze correlation without improving choice alignment, and a single average gaze map per image position raises it by a similar amount. Matching human choices, or even human gaze patterns, is therefore not sufficient evidence of human-aligned cross-modal processing.

In regulated industries like consumer finance, seemingly harmless user queries can exploit large language model vulnerabilities, triggering safety failures and pushing responses dangerously close to policy limits. Existing automated red-teaming methods trade off attack effectiveness against generation cost, while treating coverage, severity, and diversity as incidental rather than joint objectives. We introduce FinRT, a structured framework that builds reusable adversarial prompt generators from adaptive red-teaming strategies. Across the six victim models in consumer finance, FinRT substantially outperforms adaptive search baselines while amortizing target-facing attack generation into a reusable generator. FinRT nearly doubles the attack success rate over the adaptive baseline Rainbow Teaming (32.9% vs. 17.2%), increases maximum adversarial severity by 33%, and preserves comparable intra-policy-domain semantic diversity to iterative search methods. Our method achieves high cross-model transferability while exhibiting distinct victim-family specialization patterns.

Temporal abstraction via options can improve exploration in large environments. However, existing option discovery algorithms find subgoals that target all aspects of the state simultaneously. This state-reaching approach produces options that only apply in narrow regions of the state-space, eventually causing an explosion in the number of options that overwhelms the agent, and impedes progress on its primary task of reward maximization. We introduce an algorithm that instead identifies a small, relevant subset of features for each subgoal, yielding options that generalize broadly and accelerate exploration. Our approach learns abstract, transferrable options and achieves rapid exploration in three sparse-reward, image-based domains, including the Atari game MontezumasRevenge.

Sequential Model-Based Optimization (SMBO) traditionally relies on Bayesian or ensembling surrogates for uncertainty quantification. While historically treated as fully data-driven, SMBO increasingly integrates external domain expertise to accelerate discovery. To overcome the opaque guidance and diminished integration fidelity of standard acquisition re-weighting, Probabilistic Circuits (PCs) have emerged as a generative surrogate alternative, enabling direct knowledge injection via conditional sampling. However, these generative routines lack the formal exploration-exploitation semantics required for rigorous optimization. We introduce the Distributional Conformance Score (DisCo), a novel metric that unifies the flexibility and efficiency of PCs with a formal uncertainty framework. DisCo provides a bounded, $[0, 1]$-normalized measure of model "surprise" that (1) recovers properties comparable to kernel-based uncertainty known from, e.g., Gaussian Processes, while maintaining linear-time inference, and (2) enables accurate assessment of conformance of external knowledge w.r.t. model evidence. We then present DisCoMBO, a framework leveraging these properties for robust, knowledge-aware optimization. We prove that DisCoMBO is a zero-regret algorithm and demonstrate its effectiveness across diverse benchmarks from AutoML, material optimization, and wind park optimization.

World-Action Models (WAMs) improve robotic manipulation by conditioning action generation on predicted future observations, but future prediction adds further inference overhead to already expensive iterative action generation. Action chunking can amortize this cost over multiple actions, yet performance degrades over long execution horizons because later actions remain conditioned on stale observations. We introduce STAIRCASE POLICY, a streaming inference and training framework that turns a flow-matching VLA into a JEPA-style WAM and partitions a large action chunk into sub-chunks at staggered denoising stages. Near-term actions are executed as soon as they become available, while later actions continue to be refined. At each sub-chunk boundary, the future latent is re-predicted from the latest observation and used to update all unexecuted actions, enabling long-horizon execution without repeated full policy inference. The resulting future-prediction error can further serve as a signal for adaptive chunking. S-WAM achieves 97.7% on LIBERO and 87.9% on LIBERO-Plus, and improves performance across multiple policy backbones and real-robot tasks. It reaches 292.7 executed actions per second, $3.62\times$ the throughput of conventional execution at comparable accuracy, while reducing time-to-first-action from 123.6 to 73.3 ms. With additional inference optimizations, throughput further increases to 642.9 actions per second.

Truncus arteriosus (TA) is a rare, severe congenital heart disease in which the two main arteries exiting the heart fail to separate in utero resulting in one truncus and truncal valve, carrying mixed oxygenated and deoxygenated blood. About 25% of patients have a quadricuspid valve, which is prone to regurgitation and re-intervention. Despite its relevance for valve performance and mixing, fluid mechanics in TA are poorly understood. Patient-specific fluid-structure interaction simulations were performed based on CT imaging before and after TA repair. The quadricuspid valve was constructed using elasticity-based design with the patient's free edge length and geometric height extracted from echocardiography. Interaction between blood and valve was simulated with the Immersed Boundary Method. Boundary conditions were tuned to the patient's data. Mixing of oxygenated and deoxygenated blood and streaming were assessed via Lagrangian Coherent Structures (LCS) and Lagrangian Particle Tracing (LPT). Low pressures, forward flow and streamwise vortices in the one-sided pulmonary arteries (PAs) throughout the cardiac cycle affected leaflet motion, leading to asymmetric closure and regurgitation. Holodiastolic aortic flow reversal supplied PA flow and the regurgitant jet. LCS and LPT indicated favorable streaming of oxygenated blood from the LV to the aorta and deoxygenated blood from the RV to the PAs. After truncal surgery, normal hemodynamics were restored. This is the first study of fluid mechanics of TA. Using qualitative and quantitative flow analysis, we identified disrupted preoperative hemodynamics caused by one-sided PAs and showed how normal hemodynamics were re-established after repair. Favorable streaming was demonstrated aligned with patient reports. Thus, favorable streaming is plausible in total mixing lesions and patient-specific modeling may aid in its detection.

The convergence of large language models, materials-specific foundation models, and agentic artificial intelligence is reshaping the paradigm of computational materials discovery. While high-throughput computation, automated workflows, and data-driven modeling have greatly expanded the scale of materials exploration, the core scientific decision-making loop remains largely human-directed. Agentic AI introduces the possibility of systems that can autonomously reason about materials objectives, execute simulations, and refine strategies. However, the rapid emergence of such systems has created a critical need for a unified and operational framework to define, evaluate, and guide scientific autonomy in computational materials discovery. In this Perspective, we propose the Computational Materials Agent Autonomy Level (CMA-AL) framework, a hierarchical taxonomy defining six levels of autonomous agency in computational materials science: scripted excecutor, LLM-assisted operator, adaptive explorer, experiment-ready modeler, agentic digital twin, and self-extending intelligence. We further map emerging agentic systems onto the framework and identify key scientific and technological challenges toward higher autonomy. CMA-AL provides a common language for characterizing agentic computational materials discovery, evaluating the maturity of emerging systems, and guiding their evolution toward increasingly autonomous materials discovery.

Massachusetts recalibrated its Clean Peak Standard (CPS) in 2026 by lowering the minimum standards and expanding the Near-Term Resource Multiplier for qualifying storage. This paper evaluates the change using a three-zone, hourly capacity expansion model for 2026–2030. Both scenarios achieve full CPS compliance, but the post-May scenario reduces new battery additions by 53.1% and modeled system cost by 3.7%. CPS-eligible discharge declines by 30.4%, while Clean Peak Energy Certificate production declines by only 9.5%. Fossil generation during CPS hours increases by 11.7%, whereas modeled emissions remain nearly unchanged because the Regional Greenhouse Gas Initiative cap binds. The recalibration therefore lowers the modeled storage capacity requirement while weakening physical clean-peak performance.

Let $p$ and $\ell$ be distinct odd primes. For a finite extension $F/\mathbb{Q}_p$, the local Langlands correspondence states that there is a canonical bijection between irreducible, smooth representations of $\text{GL}_n(F)$ and $n$-dimensional, $\Phi$-semisimple Weil–Deligne representations of the Weil group $W_F$. Given two $2$-dimensional, semisimple, continuous representations $\rho_1, \rho_2$ of $W_F$ with images in $\text{GL}_2(\mathcal{O}_K)$, where $K/\mathbb{Q}_\ell$ is a finite extension with maximal ideal $\lambda \subset \mathcal{O}_K$, a natural question to ask is when their mod $\lambda$ reductions $\overline{\rho}_1$ and $\overline{\rho}_2$ are isomorphic. In this paper, we give a complete characterization of when the reductions are isomorphic. For this, we first give a description of when the reductions of these representations are decomposable or irreducible, utilizing the correspondence between continuous $\ell$-adic representations of $W_F$ and $\ell$-adic Weil–Deligne representations. We conclude with some examples which arise in the modular method for solving generalized Fermat equations.