Physical AI systems, including robots, autonomous vehicles, embodied agents and edge copilots, often run a different inference workload from cloud LLM serving: single-stream, batch-1 autoregressive decode, where one robot, camera feed or user session waits on the next token. This workload is usually described as memory-bandwidth-bound. Each decode step streams model weights and the active KV cache, so latency should scale with peak HBM bandwidth. We show that this account is true but incomplete. We measure batch-1 decode for three 7 to 8B-class GQA transformers across four NVIDIA GPUs: H100 SXM5, A100-80GB SXM4, L40S and L4. We evaluate context lengths from 2048 to 16384, producing 44 valid cells under a controlled bf16 SDPA setup. The achieved fraction of peak HBM bandwidth falls as peak bandwidth rises. On the headline Qwen-2.5-7B ctx=2048 cell, an L4 reaches roughly 81 percent of its analytic memory floor, while an H100 reaches only 27 percent. Physical-AI decode is memory-dominated, but faster memory does not translate into proportional latency gains. We test the missing term with a CUDA Graphs A/B experiment. On H100 at ctx=2048, CUDA Graphs improves decode latency by 1.259x across N=10 fresh sessions, with a 95 percent bootstrap confidence interval of 1.253 to 1.267. On L4, the same intervention gives only 1.028x. This isolates a launch-side overhead that becomes visible on fast GPUs but remains mostly hidden on slower, bandwidth-bound GPUs. The deployment implication is that memory savings matter only when the runtime realises them. On L4, bf16 decode sits close to the memory floor, but common quantised paths do not recover the expected 4x weight-traffic reduction: bnb-nf4 reaches 59.36 ms/step and AutoAWQ+Marlin reaches 45.24 ms/step from a 62.32 ms bf16 baseline. GPTQ+ExLlamaV2, with Ada-tuned int4 kernels, reaches 17.36 ms/step.

We study true self-avoiding walk (TSAW) as a mechanism for improving empirical integral estimation via Markov chain Monte Carlo (MCMC). We consider finite-state adaptive sampling dynamics associated with an irreducible Markov kernel $P$ on a finite set, with stationary distribution $π$, in which the transition probabilities are penalized according to empirical overuse. Our main result is that the empirical occupation counts $L_t(i)$ and transition counts $N_t(i,j)$ of the resulting TSAW-based walk satisfy \[ L_t(i)-tπ_i = O(\sqrt{\log t}) \quad\text{and}\quad N_t(i,j)-tπ_iP_{ij}=O(\sqrt{\log t}) \qquad\text{almost surely} \] for every state $i$ and every edge $(i,j)$ with $P_{ij}>0$. Consequently, for every bounded function $f:V\to\mathbb R$, the error of our integral estimator converges as \[ \left|\frac1t\sum_{s=0}^{t-1} f(X_s)-\sum_{i\in V}π_i f(i)\right| = O\left(\frac{\sqrt{\log t}}{t}\right) \qquad\text{almost surely}. \] These results show that, in contrast with the usual $t^{-1/2}$ error scaling for empirical averages under standard random-walk-based methods, TSAW-based estimator yields empirical integral errors of order $O(\sqrt{\log t}/t)$ almost surely, thereby achieving a substantially sharper dependence on the sample size $t$.

We present improved bounds for estimating discrete probability distributions under the $\ell_\infty$ norm. These include minimax bounds in expectation and high-probability tail bounds. We resolve some of the open questions posed in Kontorovich and Painsky (JMLR, 2025) -- including a fully empirical version of the tightest risk bound they presented and identifying the form of the worst-case extremal distribution. Encouraging empirical results are reported as well.

Reinforcement learning with verifiable rewards (RLVR) trains language models using programmatically checkable signals such as unit-test outcomes, enabling direct optimization for functional correctness in code generation. We conduct an empirical study of RLVR for Python code generation on the MBPP benchmark using two small models (Qwen3-0.6B and Llama3.2-1B) with LoRA fine-tuning. Across multiple reward formulations such as: unit-test-only rewards, static-analysis-only shaping via the Ruff linter, and a combined reward, we compare group-based policy optimization variants (GRPO and GSPO) and evaluate both functional correctness and behavioral diagnostics. In our experimental setting, RLVR improves pass@1 on MBPP test by up to 13 percentage points under proposed combined reward configuration. However, we find that reward shaping can induce systematic behavioral shifts: using only static-analysis penalties may bias the policy toward shorter completions that reduce lint errors without reliably improving functional correctness. In contrast, combined rewards mitigate this degeneration and yield more stable trade-offs between correctness and style constraints. Overall, our results highlight that RLVR effectiveness for code generation is highly sensitive to reward design and optimization granularity, and that diagnostics beyond pass@1, including generation length, Ruff severity profiles, and execution error types are useful for identifying failure modes.

We study privately estimating the sum of $n$ user-held values in the presence of an honest-but-curious server. This motivates requiring privacy not only at data release but also throughout server-side computation. We therefore adopt the local (pure) differential privacy model, in which each user transmits a noise-perturbed value. It is well known that independent local noise typically incurs a substantial utility loss compared to the centralized model, where noise is added only after aggregation. We show that this gap is not fundamental. By carefully designing correlations among the locally added noise variables, we construct $\varepsilon$-DP mechanisms whose estimation cost matches the optimal cost achievable in the centralized setting, up to an arbitrarily small error.

Tool-augmented LLM agents are vulnerable to prompt injection: a third party who controls part of the agent's context can plant instructions that the agent then executes as if they came from the user. Current evaluations report a single attack success rate per model on one channel, the tool output and treat that number as the model's vulnerability. But tool descriptions, which the agent reads at every turn before any tool is called, are themselves an injection surface that the attacker can choose instead. We hold the injection payload byte-identical and deliver it through both surfaces across 13 LLMs from six families and four task suites. The same bytes invert in success rate across models: GPT-4.1 is 96 percent vulnerable on tool outputs but only 4 percent on tool descriptions, while GEMINI-3-FLASH shows the mirror pattern at 20 percent and 98 percent. A variance decomposition over 6,830 attempts attributes 0 percent of the variation in attack outcomes to the surface alone, while the model-surface interaction accounts for 16.7 percent. Vulnerability is a property of the pairing, not the channel. The Adaptive Attack Rate, defined as the per-cell maximum over surfaces, exceeds the strongest fixed-surface baseline by +9.1 percentage points on average. Standard prompt-level defenses inherit the same blindspot, reducing tool-output ASR to 10-18 percent while leaving the description channel above 54 percent. Both attack and defense evaluation must report per-surface vulnerability.

Generative machine learning has become an essential tool in theoretical and experimental physics, especially in the context of fast surrogates and density estimators. In this work, we first introduce the underlying framework of modern generative networks and then discuss challenges in quantifying their accuracy, precision, and statistical power.

Finding symmetries is crucial for understanding physical models. In this work, we present an optimization framework that searches Pauli symmetries of Hamiltonians, merging the fields of machine learning with automated symmetry finding. Built on a Set-Transformer architecture, our framework uses self-attention to encode the pairwise and higher-order correlations among the Pauli-Strings. The relations are then decoded as a candidate, which is further optimized with a custom commutation-based objective, and mapped to a symmetry of the input Hamiltonian. We apply our method to random Pauli Hamiltonians, periodic one and two dimensional transverse-field Ising model and the Toric code. We show that for physical Hamiltonians (Ising and Toric), our framework succeeds with near-deterministic probability while providing substantial advantage compared to state-of-the-art strategies. For random Pauli Hamiltonians, we estimate the required computational resources, specifically the number of parallel starts and the number of GPUs, to find a symmetry with high success probability under fixed design specifications.

Recent research demonstrating AI systems exhibiting deception and shutdown resistance suggests that AI loss of control (LOC) is an urgent policy concern , yet current literature focuses almost exclusively on alignment and prevention. To address this gap, this paper introduces a foundational framework and taxonomy for managing catastrophic AI LOC incidents. The taxonomy's first level distinguishes between scenarios where regaining control is 'extremely costly' versus 'impossible'. While impossible scenarios demand immediate resilience investments to fundamentally restrict an AI's attack surface , extremely costly scenarios require active incident management via Containment and Threat Neutralization. The framework further categorizes these manageable events into accidental LOC (requiring automated circuit-breaker responses) and adversarial LOC (requiring graduated escalatory measures). By mapping three severity classes to specific scenario matrices, this paper provides a concrete, proportional guide for managing unprecedented AI risks.

The aquaculture industry needs to address several challenges to secure sustainable seafood production that can serve an increasing global demand. One major challenge is to ensure good fish health and acceptable welfare during production since the improvement of fish welfare is of vital importance in current and future production systems. In this study, this is addressed by developing and implementing methods to identify fish behaviors in response to intrusive objects both on individual and on a group basis. A novel approach for detecting, tracking, and estimating the 3D position of individual fish has thus been developed, and specifically designed to track the caudal fins of farmed fish in industrial sea cages. The tracking data was subjected to a novel stereo-vision method adapted to estimate fish positions, velocities, accelerations, and turning and pitch angles. Datasets obtained from industrial-scale fish farms were then analyzed to identify the impact of structures of varying shapes, sizes, and colors on fish behavior. The method was trained using manually labeled caudal fins, and used YOLOv8 with ByteTrack as an object detector and tracker, SuperGlue for matching detections in the left and right frames, and triangulation to reconstruct the 3D positions of the fish. Different image pre-processing and augmentation methods for enhancing object detection accuracy were tested and their performance compared, while RAFT-Stereo was tested for depth estimation purposes. The obtained results both validate the method's performance against previous research efforts, and demonstrate the novelty and potential of this method in providing more insight into behavioral dynamics in sea-cages.

3D Gaussian Splatting (3DGS) is a promising neural scene representation for real-time rendering, but trained models often suffer from large memory footprints, limiting deployment on less powerful devices. Existing compression techniques often lead to architectures with several additional trainable parameters. While achieving outstanding compression ratios, they introduce noticeable drops in image quality. In this work, we introduce the first dictionary-learning-based compression framework for 3DGS. The proposed post-training compression pipeline can be deployed in virtually any 3DGS model without the need for re-training or modifications to existing 3DGS models. Our compression framework is straightforward to implement, yet provides significant compression capabilities, preserves image quality, and improves real-time rendering performance. Across 13 benchmark scenes, our approach achieves an average compression ratio of 3.95x, 3.10x, and 4.55x when applied to 3DGS, 3DGS-MCMC, and PixelGS, respectively. This yields consistent rendering speedups of 23.3%, 24.3%, and 25.3%, while maintaining image quality.

This paper introduces Code Bench, a benchmark capable of evaluating Large Language Models (LLMs) concise code generation abilities in 60 programming languages. Based on code golf, a recreational programming competition focused on minimal character or byte solutions, the benchmark provides a distinctive measure of LLMs ability to produce efficient, concise code. Unlike existing benchmarks limited by fixed problem sets and language coverage, CodeGolf Bench leverages the code.golf platform to provide new problems and live human performance baselines. Evaluation of nine LLMs on Python and C++ tasks demonstrates that reasoning models significantly outperform non-reasoning models, achieving best average percentile of 70.97%. This performance gap is particularly pronounced in C++, highlighting reasoning's importance for languages with strict syntax requirements. Non-reasoning models struggle more with efficiency optimization across both languages, with best percentiles significantly lower than reasoning counterparts. CodeGolf Bench offers a dynamic framework for evaluating LLM code generation capabilities against evolving human performance on code golf.

Human reasoning has long been theorised to operate socially, not through isolated individual cognition, but through collective adversarial discourse, a framework known as the Argumentative Theory of Reasoning (ATR). Rather than relying on individual "intellectualist reasoners" as the primary vehicle for truth-seeking, ATR reconceptualises truth as an emergent property of social epistemology: the product of imperfect individual reasoning refined under the adversarial pressure of debate. This distributed method of collective intelligence has guided humanity to ever-greater epistemic heights and underpins the foundational principles of all democratic systems. This thesis breaks new ground by, for the first time, simulating ATR through the multi-agent debate (MAD) of large language models (LLMs). With rigorous empirical analysis, we demonstrate that, when correctly engineering an epistemically diverse set of models, LLM-MAD can significantly improve truth-seeking performance on questionnaire-based tasks, even when individual debate participants exhibit limited standalone performance. Furthermore, we present strong empirical evidence that this performance gain is mechanistically grounded in the central principles of ATR, suggesting that collective reasoning may be universally favourable over individualist reasoning, rather than a quirk in biology or evolution. Finally, drawing on our analysis of debate dynamics, we propose a novel benchmarking methodology that leverages LLM-MAD to measure intrinsic model properties (such as hallucination propensity) in order to compare models in ways that current static benchmarking approaches cannot support.

Pattern languages are a classical model in formal language theory and algorithmic learning theory. This note formulates the problem of computing the inclusion depth of a pattern language: the length of the longest strict inclusion chain from the universal pattern language to the language generated by a given pattern. Inclusion depth captures the mind-change complexity of pattern identification from positive data. The central open question is whether the inclusion depth ID_Sigma(p) is computable for every pattern p over every finite alphabet Sigma with at least two symbols, and whether it is computable in polynomial time. A simple conjectured formula, ID_Sigma(p) = 2|p| - #var(p) - 1, would imply a linear-time algorithm. The problem connects pattern language inclusion, combinatorics on words, language identification in the limit, and mind-change-bounded learning.

High-fidelity computational fluid dynamics is essential for aerospace design, but engineering-scale simulations of practical three-dimensional aircraft remain computationally expensive. Learning-based flow-field initialization can improve efficiency by reducing the numerical distance between the initial and converged solutions, yet existing deep learning approaches remain difficult to scale to large three-dimensional aircraft flows with multiscale regional heterogeneity. Most prior studies therefore focus on two-dimensional problems, surface quantities, integral aerodynamic coefficients, or simplified three-dimensional cases with limited grid resolution.Here we propose MHLF, a multigrid-hierarchical learning framework for accelerating engineering-scale aircraft flow simulations while preserving high-fidelity numerical accuracy. MHLF combines a topologically consistent geometric multigrid representation with a hierarchical strategy that captures regional flow heterogeneity during both prediction and subsequent CFD correction. Across three engineering-scale aircraft cases spanning Mach 0.15 to 6.0 and covering subsonic, transonic and supersonic regimes, MHLF accelerates convergence without sacrificing flow-field accuracy, achieving a 3 to 8 times efficiency improvement over conventional initialization. These results demonstrate practical full-flow-field prediction for large three-dimensional aircraft within the CFD domain and provide a foundation for data-driven acceleration of high-fidelity aircraft flow simulation.

Reservoir computing, a type of recurrent neural network, is a promising approach for temporal learning as it separates dynamic processing from the trained readout layer. However, classical Echo State Networks (ESNs) often require task-specific tuning of their architecture and hyperparameters to achieve good performance. This paper introduces EARLY (Evolutionary Algorithm for Reservoir Learning and Yielding), a framework designed to evolve both the topology and hyperparameters of multi-reservoir ESNs. Inspired by the modular organisation of the brain, EARLY encodes architectures as graph-based genomes and applies crossover, mutation, and selection to discover effective configurations. Our goal is to create both generic architectures and tasks inducing generalization. The method is evaluated on temporal learning tasks from the CogScale dataset. Results show that evolved architectures outperform those obtained with random search on several tasks and exhibit structural differences depending on task difficulty: simpler tasks yield lightweight architectures, while more complex tasks favour richer modular organisations. These findings suggest that evolutionary search can help identify reusable reservoir structures for a broader range of temporal problems. The evolved architectures are further evaluated on a cross-situational learning dataset to assess their ability to adapt to new environments.

We address the issue of global convergence in stochastic continuous optimization. For that purpose, we formulate the Canonical Evolutionary Strategy (CES) as a controlled mathematical framework to analyze global convergence in evolutionary algorithms via the semiclassical limit of a Schr{ö}dinger-type replicator-mutator equation. We provide a rigorous hierarchy from a discrete individual-based dynamics to a deterministic mean-field limit, demonstrating that global convergence is governed by the principal eigenfunction of the underlying operator. This property, defined as Geometric Selection, naturally prioritizes robust, flat optima over narrow local traps, offering a mathematical justification for the ''survival of the flattest'' phenomenon. Moreover, unlike consensus-driven methods that are prone to premature variance collapse when the global minimizer resides outside the initial support, the replicator-mutator dynamics of CES facilitate intrinsic mass transport. High-dimensional benchmarks (d = 30) confirm this advantage, showing that CES achieves lower residual errors in shifted initialization scenarios where standard consensus-driven and gradient-based methods fail to migrate effectively. By shifting the focus from point-wise consensus to spectral concentration, our framework provides a robust theoretical foundation for global convergence in Evolution Strategies (ES) without the need for additional numerical heuristics.

Surrogate Safety Measures (SSMs) are extensively utilised in the evaluation of traffic risk in automated driving contexts. However, the majority of SSM-based evaluations employ fixed thresholds that fail to capture the human response to sustained borderline conditions or the reaction to brief, high-risk peaks. The present work proposes a biologically inspired reinterpretation of SSM thresholds. This is modelled as spiking thresholds of leaky integrate-and-fire (LIF) neurons, with multiple SSM inputs combined into a spiking neural network (SNN). The SNN is trained to emit spikes that are aligned with human braking onsets. The training data was recorded in a controlled car-following experiment using the 3D-CoAutoSim platform with CARLA/Unreal and a 6-DOF motion platform, where induced critical events were generated. The results demonstrate that the learned spiking activity qualitatively aligns with braking behaviour across scenarios and captures reactions that are not consistently explained by threshold crossings alone. Analysis across participants further indicates that learned input thresholds remain relatively consistent, while learned decay factors encode different temporal sensitivities for the SSMs. The findings of this study indicate that spiking dynamics may serve as a mechanism to facilitate the convergence of objective SSMs with subjective human safety perception.

Recent Singing Voice Synthesis (SVS) advances enable highly realistic but potentially malicious AI covers, making singing voice deepfake detection (SVDD) crucial. Self-Supervised Learning (SSL)-based detectors achieve state-of-the-art performance by fine-tuning speech SSL backbones to capture singing-specific spoof artifacts. Existing adversarial attacks often fail against SSL-SVDD, creating a false impression of inherent robustness. We reveal this stems from two challenges. First, at the objective level, attacks optimize cross-entropy on local surrogates, crossing surrogate-specific boundaries rather than suppressing shared spoof evidence. Second, at the method level, attacks follow the surrogate's dominant gradient direction. In SSL-SVDD, this aligns with fine-tuned artifact-sensitive directions, limiting transferability to unseen detectors - a geometric failure we term the Linearity Trap. To properly evaluate robustness, we propose MARS (Meta-Adversarial Regression of Semantics), a transfer-based black-box framework tailored to SSL-SVDD. Structurally, MARS shifts to hypothesis-evidence manipulation by constructing a natural semantic anchor from the pre-trained SSL space and an artifact anchor from the fine-tuned space. Algorithmically, MARS escapes the Linearity Trap via bi-level optimization: the inner stage induces tangential exploration, while the outer stage guides the audio toward the natural semantic manifold. Experiments on the CtrSVDD benchmark show MARS improves Attack Success Rate (ASR) in in-distribution transfer (13%), out-of-distribution transfer (10%), and cross-task evaluation (36%), highlighting the urgent need for robust SVDD systems.

Radio-frequency (RF) fingerprinting identifies wire-less transmitters using hardware-induced imperfections present in baseband I/Q signals. However, deep learning models often degrade under receiver and channel distribution shifts, particularly as transmitter populations grow. This work proposes the Hamiltonian Transformer, a physics-informed attention architecture that enforces norm preserving value dynamics within each attention head using a learned skew-symmetric generator and a Störmer-Verlet leapfrog integration step. An additional phase-increment embedding exposes oscillator dynamics at the input layer. All experiments use non-equalized raw I/Q signals from the WiSig dataset under four protocols: same-day classification, cross-receiver generalisation, cross-day generalisation, and transmitter scaling up to 150 devices. The Hamiltonian Transformer achieves 99.12% accuracy under same-day conditions and 61.64% at 150 transmitters, consistently outperforming CNN and Transformer baselines across all scale points. A controlled ablation study identifies norm-preservation in the value update as the primary inductive bias driving the scaling advantage, with the phase increment embedding providing the single largest per-component improvement. These results indicate that embedding physics-informed structural priors into attention mechanisms is an effective approach to large-scale transmitter identification on raw wireless signals.

Regime shifts in financial markets reorganise the joint dynamics of asset prices and macro variables, breaking any single-regime calibration. They are nonetheless difficult to detect reliably because the data signal is noisy and heavily multicollinear, while the contemporaneous text that announces them is unstructured. Standard regime shift detection methods rely solely on structured time-series data and ignore policy communications, even though these texts often signal shifts before they materialise in observed prices. We propose a text-enhanced regime shift detection pipeline that combines large language model (LLM) reasoning over central-bank communications with statistical validation on multivariate financial time series. The framework is detector-agnostic: text-proposed candidates are validated using a bootstrap likelihood-ratio test on a vector autoregression (VAR), while data-driven candidates from arbitrary regime detectors are ratified through a lenient LLM text check. We evaluate the framework on 2010-2024 FOMC minutes paired with a 14-variable U.S. Treasury and macroeconomic panel, using four interchangeable data-driven detectors. The proposed pipeline achieves F1 = 0.82 against a verified anchor list of monetary-policy regime shifts, with same-day modal detection latency and consistently stronger performance than pure data-driven baselines. The results demonstrate that combining unstructured policy text with statistical structural-break detection improves the robustness and interpretability of regime shift identification in financial markets.

Spiking neural networks (SNNs) hold promise for demonstrating superior learning and representation capabilities in deep models. Given the tremendous success of ResNet in deep learning, it would naturally follow to train deep SNNs with residual learning. However, existing residual structures for constructing deep SNNs still present challenges of spike redundancy or information loss, as well as redundant learning. In the present study, we first aim to address issues of relative spike redundancy in identity mapping and information loss in non-identity mapping. To this end, we propose an OR-ADD (OA) shortcut connection to merge output spikes/currents from two branches in the residual structure. Furthermore, to mitigate redundant learning in the backbone branch of the residual structure, we introduce the concept of XOR meta-residuals, i.e., selecting pre-learning residuals using the Exclusive-OR (XOR) operation for the backbone branch. Finally, by integrating the OA shortcut and XOR meta-residuals, we devise the XOR residual block and further construct XOResNet with varying depths based on this block. Extensive experiments on four datasets, Fashion-MNIST, CIFAR-10, CIFAR-100, and miniImageNet, show that the proposed XOResNet outperforms existing state-of-the-art deep SNNs optimized via gradient descent. These results validate the effectiveness of our OA shortcut and XOR meta-residual components in overcoming fundamental limitations of residual learning in SNNs, providing new architectural insights for building high-performance neuromorphic systems.

Spiking Neural Networks (SNNs) offer compelling energy efficiency on neuromorphic hardware, yet their training remains challenging because the discrete spike threshold is non-differentiable. Surrogate-gradient methods sidestep this by approximating the derivative, but they impose backpropagation infrastructure that is incompatible with on-chip learning. Evolution Strategies (\es) are a natural gradient-free alternative, yet their computational cost scales with the number of parameters, making them impractical for large weight matrices. We present a method for training SNNs using EGGROLL, a low-rank factorisation of ES perturbations that reduces per-generation memory from $\mathcal{O}(mn)$ to $\mathcal{O}(r(m{+}n))$. Combining EGGROLL with a Leaky Integrate-and-Fire SNN on N-MNIST, we demonstrate that gradient-free training achieves 79.21% test accuracy while reducing per-generation wall-clock time by 2.23$\times$ relative to full-rank ES. Our results demonstrate EGGROLL is viable for SNN training, with a clear accuracy-speed tradeoff, compatible with training on neuromorphic hardware without surrogate gradients.

Generating high-performance GPU kernels remains challenging due to the need for both correctness and hardware-aware optimization. While large language models (LLMs) show promise in code generation, they often fail to produce kernels that are both correct and efficient. We propose Kernel Foundry, a diagnosis-driven evolutionary framework for automatic GPU kernel optimization. Our method combines expert-guided, retrieval-augmented initialization with a multi-island evolutionary search, where candidate kernels are iteratively refined using structured diagnostic feedback. A centralized experience library accumulates reusable optimization knowledge to guide subsequent evolution, while explicit mechanisms prevent cheating behaviors that bypass kernel-level computation. Experiments on KernelBench show that our method consistently improves both correctness and performance over strong baselines, achieving up to 100% correctness on Level~2.

GUI agents rely on screenshots to infer intent and operate across applications, but these screenshots often contain private messages, medical records, payment credentials, and workplace-specific workflows. Privacy decisions in this setting depend on task, recipient, application state, and user role, yet static PII detectors miss these boundaries and cloud-side VLM reasoning can upload the raw screen before deciding what should be protected. We present MaskClaw, an edge-side privacy arbitrator for GUI agents. MaskClaw extracts local visual evidence, retrieves user- and task-specific policy memory, and decides Allow, Mask, or Ask before raw screenshots leave a trusted user- or organization-controlled environment. In five designed skill-evolution scenarios, it turns corrections, cancellations, and edits into reusable privacy skills checked by a sandbox gate. We introduce P-GUI-Evo, a benchmark built from real UI patterns, reconstructed HTML screens, and sanitized labels. Experiments show that pattern matching, cloud reasoning, and routing alone tend to over-confirm, over-mask, or expose raw screenshots under the same protocol. The artifact is available at https://github.com/Theodora-Y/MaskClaw.

Logical operations are essential for quantum computation within quantum error-correcting codes. However, discovering their physical realizations is challenging, especially for non-additive codes that lack a stabilizer description. We present a general learning-based framework that, given only an encoding circuit, constructs physical implementations of logical operations while enforcing structural properties such as transversality or shallow depth. Our approach is validated by rediscovering known logical operations of standard stabilizer codes. We then extend it to a co-design procedure, dubbed variational early fault-tolerant quantum computing (VarEFTQC), which tailors non-additive encodings to a given noise model and enforces desired logical gate sets, such as transversal IQP-type families or low-depth universal sets. A software library implements the complete learning pipeline, including loss-function variants, ansatz families, and optimization routines. Together, these results position VarEFTQC as a practical tool for discovering hardware-adapted logical gadgets for early fault-tolerant quantum computing.

We study efficient differentially private algorithms for estimating monotone statistics, i.e., statistics that are monotone under the addition of new observations. The starting point for our investigation is subsample-and-aggregate: a classical paradigm that partitions the dataset into blocks, estimates the statistic on each block, and then privately aggregates the estimates. While practical and generically applicable, this approach is quite data-hungry. We improve upon this framework for the class of monotone statistics -- compared to subsample-and-aggregate, our algorithms save a factor of $t$ in sample complexity and pay a factor of $e^t$ in running time, where $t>0$ is a tunable parameter. We complement our results with a query-complexity lower bound, showing that our algorithms are essentially optimal for this task. As an application, we obtain improved results for private eigenvalue estimation, private loss estimation, and privately estimating a single parameter of a high-dimensional model, e.g., in linear regression.

Not all clinically relevant adverse effects are structurally inferable from molecular graphs - regardless of model quality or architectural complexity. This study introduces an operational taxonomy of the structural information limits that prevent structure-based toxicity prediction, independent of the learning algorithm employed. Graph Neural Networks (GNNs) have emerged as a natural approach for molecular toxicity prediction, operating directly on atomic connectivity without the information loss inherent to fixed-length fingerprints. However, the fraction of a drug's known pharmacological profile that is actually inferable from molecular structure remains systematically underexplored. A systematic case study using acetylsalicylic acid (ASA, Aspirin) - one of the most comprehensively characterized drugs in pharmacology - serves as model compound. A Message Passing Neural Network (MPNN) is trained on the Tox21 benchmark and GNNExplainer is applied to characterize atom-level attribution. Results indicate that molecular structure explains approximately 45% (5/11) of known ASA adverse effects. A four-category Gap Taxonomy (GAP-1 through GAP-4) is introduced distinguishing between principally non-encodable effects, data gaps arising from Missing Not At Random (MNAR) mechanisms, assay panel mismatches, and representation errors. The MNAR gap is empirically quantified via a systematic ChEMBL query (42 documented assays, 0 retrievable bioactivity entries). An attention pooling experiment localizes the representation error to the MPNN message passing layers rather than the aggregation step. The Gap Taxonomy has direct implications for drug safety signal detection and regulatory frameworks including Good Pharmacovigilance Practice (GVP) guidelines and New Approach Methodologies (NAMs). Structural limits identified are confirmed in a companion DDI ablation study.

Learning-to-Defer (L2D) methods route each query either to a predictive model or to external experts. While existing work studies this problem in batch settings, real-world deployments require handling streaming data, changing expert availability, and shifting expert distribution. We introduce the first online L2D algorithm for multiclass classification with bandit feedback and a dynamically varying pool of experts. Our method achieves regret guarantees of $O((n+n_e)T^{2/3})$ in general and $O((n+n_e)\sqrt{T})$ under a low-noise condition, where $T$ is the time horizon, $n$ is the number of labels, and $n_e$ is the number of distinct experts observed across rounds. The analysis builds on novel $\mathcal{H}$-consistency bounds for the online framework, combined with first-order methods for online convex optimization. Experiments on synthetic and real-world datasets demonstrate that our approach effectively extends standard Learning-to-Defer to settings with varying expert availability and reliability.

A learning-to-defer (L2D) system decides, for each input, whether to predict on its own or to hand it to one of several available experts. The very well established recipe trains classifier and router jointly by treating the $K$ classes and $J$ experts as competing actions in one shared $(K{+}J)$-action geometry. Subsequent work has proposed a series of incremental fixes within this geometry; we show that each still suffers, to varying severity, from an optimization-level pathology (target distortion, gradient amplification, winner-take-all starvation, set-mass collapse, or class-expert coupling) even under statistical consistency. We step outside the augmented-action family entirely and propose a decoupled surrogate: a softmax classifier head and an independent sigmoid head per expert, mirroring the two natural objects of the problem. We show that per-sample updates are then coordinatewise and the class-expert Hessian block is identically zero, and prove an excess-risk bound with calibration constant $\max\{2\sqrt{2},\sqrt{2J/λ}\}$ -- to our knowledge the first multi-expert L2D guarantee whose constant does not grow with the expert pool when the per-expert weight is held fixed. On controlled synthetic studies and on CIFAR-10, CIFAR-10H, and Covertype, it is the only method in our comparison that remains stable as the expert pool grows, preserves rare specialists, and improves over a standalone classifier on every real-data benchmark.