Papers

We study the number of exact matrix-vector queries needed to sample a centered $d$-dimensional Gaussian with condition number at most $\kappa$ to total variation error $\varepsilon$. The best known algorithm uses $O(\min\{d,\sqrt{\kappa}\log(\frac{d}{\varepsilon})\})$ queries, while the known lower bounds do not grow as $\varepsilon\to0$ and match this bound only for $\kappa\le d^{1/5-\delta}$ at constant accuracy. We introduce a construction based on a hidden determinant sign and prove a lower bound of $\Omega\!\left(\min\left\{d,\sqrt{\kappa}\log\left(1+\frac{d}{\sqrt{\kappa}\,\varepsilon}\right)\right\}\right)$ uniformly for all $d\ge1$, $\kappa\ge4$, and $0<\varepsilon\le\tfrac{1}{16}$. This bound matches the known upper bound up to constant factors whenever $\kappa\le d^{2-\delta}$ for fixed $\delta\in(0,2)$, or $\varepsilon\le d^{-c}$ for fixed $c>0$. We also extend a recent fixed-spectrum construction with a hidden eigenbasis from $\kappa=2$ to $2\le\kappa\le d^{1-\delta}$, for any fixed $\delta\in(0,1)$. This gives an $\Omega_\delta(\sqrt{\kappa}\log d)$ lower bound at constant accuracy that holds even when the spectrum of the precision matrix is known to the algorithm.

Certifiable optimization provides global optimality guarantees for non-convex state estimation in robotics. However, under severe physical sensing degeneracy, convex Semidefinite Programming relaxations solved via Interior-Point Methods suffer from rank expansion, rendering direct semidefinite solving and spectral rounding algebraically invalid. In this paper, we establish a universal candidate-driven dual certification framework that resolves this challenge without incurring rank expansion. First, from an epigraph perspective, we prove that while sensing degeneracy expands the primal optimal set into a flat equi-cost face, cost tightness and multiplier recovery remain strictly intact, enabling global verification without direct semidefinite solving. Second, we propose an observability-aware local search using solution remapping, coupled with a closed-form dual multiplier recovery scheme that executes in sub-microsecond latency. Third, we derive a model-specific adaptive dual safety threshold parameterized by the minimum non-zero eigenvalue of the observable Hessian and uneradicated residual leakage, preventing false rejections of optimal candidates under weak observability. We validate our universal framework with robust registration under degenerate measurements.

In reinforcement learning, an agent must explore an unknown environment to learn how to make good decisions. Unlike methods that can freely sample any state–action pair, an agent learning from a single trajectory must spend additional time navigating between states. It is therefore natural to expect that learning from a single trajectory requires more experience. In this paper, we show that this additional cost is not always necessary. For average-reward reinforcement learning, we develop an algorithm that learns where to collect information while ensuring that the required samples can be obtained efficiently. We establish an upper bound on the number of interactions needed to identify a near-optimal policy. Furthermore, we prove a matching information theoretic lower bound, showing that, for a broad class of environments and at sufficiently high accuracy, learning from a single trajectory is no harder in the worst case than learning with unrestricted access to state action samples. Our results reveal that the additional complexity cost of navigation can be avoided by adapting information gathering requirements to how the environment provides information.

This paper develops a framework for quantifying operational flexibility in radial electric distribution systems under coupled active- and reactive-load uncertainty. Flexibility is defined as the largest scaling of a prescribed uncertainty set that can be accommodated while satisfying network and actuator limits under a given control architecture. We consider a one-sided Euclidean set for robust demand-growth capability, with a direct extension to bidirectional variations. Under LinDistFlow, controllable distributed resources and feeder-head exchange with the upstream grid jointly balance active-power disturbances, while local Volt/VAR feedback regulates reactive power. The resulting affine closed-loop disturbance response admits exact second-order-cone robust counterparts. For fixed Volt/VAR gains, a convex reformulation jointly optimizes the flexibility margin and active-power participation factors, including feeder-head participation. A sparse load-to-generator feedforward map further captures communication constraints. Numerical studies on the Iowa-240 feeder demonstrate how the amount of control authority and communication structure shape the achievable flexibility margin.

We study how to improve LLM reasoning at inference time without updating model weights or using external reward models. Recent reward-free power-sampling methods define the sequence-level power target $\pi_\alpha(y\mid x)\propto p_0(y\mid x)^\alpha$, where $\alpha>1$, which increases the relative probability of completions that are already likely under the base model. Existing methods use Metropolis–Hastings or Sequential Monte Carlo (SMC) to approximately sample from this target. However, two challenges arise at different levels. At the target level, full-sequence sharpening can concentrate probability on a few exact reasoning–answer trajectories, suppressing alternative paths that support the same answer. At the sampler level, finite-particle SMC can suffer from path collapse when resampling repeatedly duplicates a few high-weight prefixes. To address these challenges, we propose Decoupled Marginal Sharpening (DMS), which separately controls how strongly the model favors particular reasoning paths and how strongly it favors final answers after aggregating support across those paths. At the target level, DMS first samples from a sharpened reasoning distribution, then marginalizes the model's answer probabilities over the resulting weighted reasoning population and sharpens the induced answer marginal. Separate exponents allow mild reasoning sharpening to encourage exploration of diverse paths, followed by stronger answer-marginal sharpening to concentrate probability on answers with greater aggregate support across those paths. At the sampler level, we develop Fully Adapted Annealed SMC, which combines an annealed sequence of power targets with locally fully adapted proposals and predictive ancestor resampling. Experiments on MATH500, GPQA-Diamond, HumanEval, and AIME24/25 across three base models show that DMS achieves the highest mean pass@1 among the compared methods in all tasks.

Bias audits of multilingual language models usually compare a model's average stereotype score in English with its average in another language and read the difference as a difference in fairness. We ask how much of that difference is noise. Auditing ten open models in eight languages, we measure how far scores move when English test items are merely reworded, test every language shift item by item, and check that the test catches shifts we inject on purpose. In all 70 model-language pairs the stereotype score falls below its English level, and for Hindi, Korean and Swahili the drop clearly exceeds what rewording English produces; most of it comes from StereoSet items. Judging every translation, we find that imperfect ones drop more, but faithful translations alone drop nearly as much. Still, a lower average does not show that a model is fairer. An item's English score predicts its translated score far less well than its reworded score (correlation 0.23, 0.27 on faithful translations, vs. 0.48), while scores stay as far from neutral, models agree on translated items, and within Hindi or Spanish a translated item's score correlates with its own rewording at 0.84 and 0.82. The average cannot distinguish lower bias from a different pattern of item-level scores. We also document three common pipeline errors that silently erase bias signal, and release our code.

Real-time volumetric mapping for a planning robot is dominated by the cost of integrating and storing information at the finest resolution, yet only a small fraction of any map ever lies on or near a path the planner actually considers. We exploit this with a tiered-resolution mapping mechanism that integrates LiDAR into a single map structure at two resolutions simultaneously: fine inside a corridor around the robot's current plan, and coarse everywhere else. As opposed to traditional navigation pipelines where path planning uses the mapping result and provides no feedback, we propose a feedback loop where planning also informs mapping to dynamically integrate information at different leaf resolutions. Across two real-world teleoperated datasets, replayed and scored offline against an independent full-resolution map, the tiered maps are 13-15x smaller and require 6-8x less integration CPU time than a uniform fine map, while producing paths at a similar cost per meter as the ones generated from fine resolution maps on both scenarios. A uniform coarse map reaches the same size savings but its path quality collapses (+210% cost per meter) on the harder scenario, because it over-discretizes terrain the robot must traverse. Tiering approximates the coarse map's cheapness without its failure mode.

Pretrained vision-language-action models (VLAs) often fail midway through unseen long-horizon tasks. Recursive self-improvement (RSI) offers a remedy, but embodied RSI methods either refine the harness around a frozen executor and so cannot add a missing motion, or train the executor on persistent failures without asking whether a failure lies in how the executor is called or in what it can do. We ask when a self-improving embodied system should change the weights of its executor, and propose AdaRSI, an on-demand RSI framework for embodied agents. A coding agent first improves its skills and harness, then attributes persistent failures by retrying failing sub-steps from saved states, and fine-tunes only the sub-steps that remain unreliable, with demonstrations it collects itself and experience carried over from earlier tasks. On all 50 target tasks of RoboCasa365, with GR00T N1.5 as the executor and no human demonstrations, AdaRSI improves the VLA on all three suites and raises its success on unseen composite tasks from 4.6% to 13.8%. Compared with fine-tuning every sub-step, it reaches slightly higher success while using only 28% of the GPU time and 22% of the self-collected demonstrations, and fine-tuning without a harness trains more yet gains far less.

Opponent histories can mislead after behavioral shifts. SafeAdapt-Poker formulates budgeted adaptation under opponent shift as a dual-counterfactual sequential-control problem using only information available before each decision. At each decision the controller asks two distinct questions: what local downside a deviation from the frozen anchor would induce under matched legal continuations, and what future adaptation value would be lost by consuming part of the remaining deviation budget now. The first question is answered by a monotone strategic-risk model over deviation intensity; the second by a paired budget-state continuation model that prices current budget consumption. The controller selects among support-preserving candidate deviations by trading immediate adaptation value against these two quantities, a pathwise cumulative-KL ledger bounds total anchor-relative deviation, and a finite family of frozen controllers is screened on complete paired calibration sessions before held-out interaction. A finite-candidate perturbation bound shows that the allocation objective degrades continuously with continuation-value estimation error. Evaluation covers exact-game exploitation and safety, four opponent shifts, local-risk calibration, delayed-opportunity allocation, direct opportunity-value estimation, value-error robustness, factorial mechanism attribution, budget sensitivity, abstract-HUNL cross-play, and player/time-held-out public-history behavior. On exact Leduc the full controller attains a matched weak-pool gain of +0.0718 with a worst-opponent delta of -0.0126, raises delayed-opportunity capture from 0.6318 under myopic allocation to 0.7841 under the same session budget, and reaches +42.7 mbb/hand in abstract-HUNL cross-play; the retained public-history screen reports 0.5820 action accuracy and 0.9277 log loss versus 0.5039 and 0.9778 for a public prior.

Cultural attitudes and social norms evolve over time, yet most cultural benchmarks for large language models (LLMs) treat culture as a static attribute of a country or region. We introduce ChronoCulture, a benchmark of 10,360 temporal pairs built from longitudinal survey data across 72 countries. Each pair contains the same survey question asked in the same country in an older and a newer year, with full population response distributions as ground truth. Across six LLMs, we find that models can often predict a country's attitudes at one point in time but do not reliably capture how those attitudes change. Their predictions for different years tend to stay close together, failing to capture much of the observed change. Specifying the intended year in the prompt also offers little overall improvement in alignment with survey responses. These findings suggest that models often present a largely static picture of cultural attitudes, highlighting the need to evaluate cultural awareness across time.

Residual scalar quantization can produce highly dependent indices when successive stages reuse identical, unscaled grids. This case study examines that behavior in a frozen neural audio codec and develops an exact recoding procedure that reduces storage while preserving its decoder inputs. Analysis of the repeated-grid recurrence identifies predictable stage sequences; finite-precision checks motivate preserving every original index through explicit exception records. The model weights and reconstruction graph remain fixed. On the studied 62- and 512-file cohorts, a predictor, adaptive coding and explicit file selection reduce complete stored bytes. The final 512-file rate is 129.591 kbps including headers, 25.943% below an original-index XZ control. Original recorded CPU runs establish bitwise latent and waveform preservation; independent readers verify exact index and metadata recovery in subsequent saved-stream experiments. Later coding choices use previously studied cohorts, and fresh encoding requires generating three alternatives. The engineering result is implementation-specific storage reduction with exact index recovery; the underlying lossy reconstruction quality remains unchanged.

Generating datasets with audio plugins requires preserving how each audio observation was produced. vst-bench-ml is a Python package that samples plugin parameters, coordinates rendering through a separately installed host, and records software identities, rendering settings, and output hashes for later verification. A sealed manifest binds this evidence and supports offline integrity checks without loading the plugin. A companion repository provides a sample-free VST3 test plugin, a deterministic offline host, and additional checks linking rendering attempts to recorded outputs. On a pinned macOS arm64 and Python 3.12 stack, functional validation of the pre-release versions reproduced package and host tests. Two runs in fresh processes reproduced the same render-plan hash and all eight expected output hashes, covering four sampled parameter configurations and two fixed MIDI signals; 12 negative cases also passed. Later licensed 0.1.0 release candidates passed producer-run continuous integration and local validation. Version 0.1.1 adds public Git source repositories, package-index installation, and user documentation; a separate release check reproduced the expected plan and output hashes. These checks establish integrity and repeatability within the tested setup; they do not establish portability, scientific suitability, or perceptual quality.

Value-based Q-learning for continuous control can discretise the action space, but the resolution of the discretisation trades early exploration against final precision. Growing Q-Networks (GQN) grow the resolution uniformly from coarse to fine. We introduce Selective Growing Q-Networks (SGQN), which keeps decoupled Q-learning and replaces uniform growth with an adaptive action grid: bins are added next to bins with high estimated Q-value, bins with low Q-value are pruned, and actions are chosen epsilon-greedily. On eight DeepMind Control Suite (DMControl) tasks with 2 to 12 action dimensions, SGQN has the highest mean final return among the compared methods on four, is within 1.1% of the best method on three more, and is 3.4% below Bang-Bang MPO on Walker Run, using on average 7.5 of 9 available bins per dimension. We then test whether the adaptive grid explains these results, and find that it does not. Growth activates every candidate bin and behaves like uniform growth, and removing pruning changes the return by at most 1.1%. Two other choices matter: starting from a coarse grid, since starting with all 9 bins produces collapsed seeds, and epsilon-greedy selection, since an annealed softmax lowers the return by 23% to 100%. SGQN is insensitive to its growth and pruning hyperparameters. It also differs from the GQN and DecQN baselines in other respects, notably one gradient update per environment step instead of one per eight, which we did not control for, so we do not attribute its advantage to a single component. We present SGQN as a competitive value-based method and as an analysis of which ingredients of a growing action grid matter.

Foundation models have shown strong generalization across scientific domains, but their use in power systems remains largely task-specific. We investigate whether a pretrained tabular foundation model (TFM), TabPFN, can serve as a general-purpose neural solver for power grid steady-state analysis. We evaluate the TFM on three tasks: power flow (PF) prediction, N-k contingency screening (NCS), and voltage stability margin (VSM) prediction. We performed experiments on 118- and 2000-bus systems, comparing the TFM against task-specific neural baselines. Without retraining or fine-tuning, the TFM matches or outperforms the best task-specific model in several settings across the three tasks. For 118-bus PF prediction, a variant of the TFM achieves errors of $1\times 10^{-3}$ pu for voltage magnitudes and $0.30^\circ$ for voltage angles, outperforming all baselines. On unseen N-1 contingency and close-to-infeasible PF cases, the TFM is competitive with the DCPF approximation. For millions of unseen N-2 contingencies, the TFM achieves AUCs of 0.92 and 0.95 for branch and voltage violation screening, respectively. The TFM also attains accurate VSM prediction performance for the 2000-bus system with limited labelled data. The TFM attains strong results across the tested tasks despite pretraining without any power-system-specific dataset and requiring no fine-tuning step. These results offer a glimpse into the potential of pretrained TFMs as a neural solver for a range of power grid analysis tasks.

Character Error Rate (CER), the standard metric for OCR and vision-language models (VLMs), charges a typo as much as a wrong name. We introduce semCER, which keeps CER's alignment and denominator and charges each grapheme edit a severity from an LLM judge, with no word segmentation. Normalized CER keeps the score of a whole transcription, where it tracks readers as well as semCER or better, and semCER breaks that score down into named edits. The charges track readers' ratings edit by edit in English, Vietnamese, and Chinese and put the edits readers rate most severe near the top of a sorted list. On pairs drawn where CER and semCER disagreed, readers side with semCER more often than with normalized CER in English and Chinese, and equally often in Vietnamese. Across 16 languages, the judge charges a wrong diacritic or tone mark almost nothing in Spanish but nearly a wrong letter's severity in Telugu. Where we can count real words, a general VLM's wrong letter forms one far above chance while a classical pipeline's stays near it. We release the 61 million judged edits, the aggregated reader ratings, and the distilled scorer at https://github.com/nguyenx2t/semcer.

Safely executing long-horizon multi-arm tasks online in complex dynamic environments remains challenging. We propose a new distributed dynamic Gaussian belief propagation (DD-GBP) method that directly plans joint-space trajectories through parallel local inference. A unified Gaussian-process (GP) parameterization connects adaptive planning horizons, interpolation-time factors, and continuous execution. Each robot maintains a local tree-structured factor graph and exchanges messages asynchronously over active inter-arm links. Dynamic subnet reconstruction and event-triggered updates reduce communication and inference costs while responding to moving obstacles. We also provide theoretical analyses of local inference accuracy, anytime early stopping, triggered-replanning consistency, and GP-consistent execution. In 50 long-horizon, multi-target four-arm scenes with moving obstacles, DD-GBP achieves $92.5\%$ collision-free all-target completion and the highest success rate among six representative baselines. Scenario ablations demonstrate long-horizon safety and robustness to communication impairment and extreme conditions; success remains $86.0\%$ under $90\%$ packet loss. Execution results further show substantially suppressed extreme velocity and acceleration peaks relative to Syn-GBP. On one workstation, obstacle-free tests with 30 coexisting arms retain $99.87\%$ success and a $15.80$ ms mean warm solve, exceeding 60 Hz and outperforming the compared methods in scalability. Dual-arm transport experiments under dynamic interference demonstrate online response and practical deployability. To the best of our knowledge, these results establish a new state of the art for online long-horizon multi-arm planning in dynamic scenes.

Chemical chain-of-thought can contain hallucinated structural and property claims. We thereby introduce MolLean, a process-verifiable language that grounds chemical reasoning in an explicit molecular state. A layered Lean-based kernel checks state transitions and assertions over molecular readouts, from primitive graph edits to compound transformations. Meanwhile, a typed witness protocol separates native structural checks from externally supplied measurements. We further introduce MolLean-MCP, a stateful tool interface that exposes these transitions, checks, and readouts during generation: the model receives feedback to decide its next action, and rejected actions leave the state unchanged. MolLean supports one-pass trace execution, whole-trace refinement, and interactive reasoning through MolLean-MCP. We evaluate these three modes with direct chain-of-thought on 8,800 matched examples across 12 tasks spanning molecular generation, editing, optimization, and reactions. With Qwen3.6-27B, whole-trace refinement outperforms direct chain-of-thought on all 12 tasks. MolLean-MCP complements whole-trace refinement with stepwise verification feedback, reaching better performance in reaction tasks. Together, these results highlight an executable foundation for chemical reasoning, in which explicit chemical commitments can be checked and revised rather than remaining solely in free-form explanations.

Co-scientists and agentic scientific claim verifiers are increasingly used to assess claims across the sciences. Natural-language scientific claims, written for human experts, are often underspecified and so admit several defensible interpretations. Such verifiers usually pick one, often tacitly, and benchmarks score their verdicts against expert labels without recording the expert's interpretation or the verifier's in a comparable form. This leads to disagreements with expert labels that cannot be traced to the evidence or the interpretation, and to compute spent on unchecked interpretations. We introduce Claim Cards to make such interpretations explicit in the reasoning itself. A Claim Card records a verifier's interpretation as nine fields before evidence gathering, each marked stated, inferred, or default. To test whether an agent's interpretation can change its verdict, we ran a GPT-5.4 agent five times on each of 60 expert-labeled materials-science and biomedical claims with a card, without one, and with one card frozen across runs (900 runs). Over 40% of card fields were not marked stated, and a card added about a tenth to a run's GPT-5.4 tokens and time without measurably changing agreement with expert labels. The agent's own interpretations varied between runs, although within a claim this variation did not measurably track the verdict. A deliberate change to one field of the frozen card, however, making the quantifier universal or the threshold for support stricter, changed the verdict on about a third of the claims, against about one in seven expected by chance (p < 0.001 for each edit). On 24 materials-science claims run with computational tools, the stricter threshold also changed what the agent asked those tools to compute. We therefore argue that verdicts should be reported with their interpretations and that benchmarks could publish the interpretation behind each label.

The scaling relationship of compute and performance for language models (LMs) has been extensively studied; more recently, models are being designed to scale memory use, as well. In particular, adding embeddings for n-grams scales memory separately from compute use, because each pass only processes individual embedding rows, which are much smaller than the storage used for the whole embedding table. However, existing methods train embeddings from scratch, which needs expensive GPU memory for fast reading and writing. Meanwhile, the ordinary memory of the host system is cheap and underutilized by these training methods. To scale on host memory, we introduce COUNT (Counting Offline Unlocks N-gram Transformers), which builds n-gram embeddings without adding learned parameters by storing precomputed n-gram counts on host. We use the corpus frequencies to weight together the LM's own unembeddings of the next tokens most often following the current n-gram history. At a fixed compute budget, COUNT can be added as an additional input embedding for pretraining 1B-parameter LMs to improve validation loss as much as an LM with 57% more compute. If GPU memory is limited, using COUNT to let backbone parameters fill GPU memory gives better performance than splitting memory with additional learned n-gram embeddings. Beyond efficiency, COUNT allows switching in a different table of n-gram counts from a domain-specific corpus at inference time to adapt to a new task without any training.

Pairwise comparison data are a standard supervision signal in preference learning, but disagreement in such data can reflect stable latent heterogeneity, ephemeral stochasticity, or non-binary responses hidden by a forced-choice interface. We study which of these semantic distinctions are identifiable from the feedback channel itself. We introduce a nonparametric latent semantic model with three querywise quantities: aggregate directional preference $\mu_q$, persistent disagreement $\tau_q$, and non-binary mass $\alpha_q$. Under the anonymous binary channel, we prove an exact semantic fiber theorem: $\mu_q$ is point-identified, whereas $(\tau_q, \alpha_q)$ vary over an explicit feasible region compatible with the same binary law. Same-annotator repeats identify $\tau_q$ through a covariance identity, abstention-enabled feedback identifies $\alpha_q$ directly, and together they point-identify the full semantic triple. We also derive finite-sample estimators, nonasymptotic concentration bounds, and semantics-aware action-recovery guarantees, and validate the theory on theorem-matched synthetic data.

Foundation models are driving a paradigm shift in robotic intelligence, moving beyond low-level action prediction to goal-directed orchestration of data collection, model training, task planning, and real-world execution. We characterize this emerging paradigm as an agentic robots, in which agents reason over human instructions, interact with robotic environments, rectify execution errors autonomously, and dynamically adjust operational strategies. Although this field develops fast, current studies are not well connected. Researchers have no consistent view of how to combine different agentic functions into a complete robot system. This paper provides a unified overview of existing research on agentic robots. We first trace the development from robot policies to agentic systems, covering vision-language-action models, world models, and world-action models. We then clarify the definition and scope of agentic robots and organize existing research into seventeen categories across four functional domains: Model, Data, Environment, and Harness. Across these domains, we examine how agents shape model learning and robot decisions, collect and curate robot data, construct worlds, tasks, and tests for learning and evaluation, and organize capabilities for robot use. Finally, we discuss open challenges and outline directions for future research. Looking forward, we hope this work contributes to a clearer understanding of agentic robots and supports new approaches to making them more capable, reliable, and adaptive.

We study classification dynamics under small initialization and identify a pronounced non-monotone behavior of accuracy that we call the accuracy well: accuracy rises rapidly early in training, subsequently drops, and only much later recovers. We observe this phenomenon in deep linear networks, fully connected ReLU networks on MNIST, deep convolutional networks on CIFAR-10, and controlled synthetic tasks, across different losses and optimizers and in both training and test accuracy. We propose a three-stage explanation in terms of target-driven organization, a degeneracy bottleneck, and subsequent recruitment and recovery. We develop a rigorous theory of these dynamics for shallow and deep linear networks, showing in particular that high accuracy can emerge while the predictor and the change in loss remain asymptotically vanishing, before continued optimization causes accuracy to deteriorate. For a simple synthetic classification task, we prove an extreme instance in which accuracy asymptotically rises to 100%, collapses to 0, and later recovers permanently to 100%. Our results show that classification accuracy can reveal aspects of learning dynamics that are not captured by the training loss alone.

Mid-2026 reporting on autonomous AI evaluation incidents claimed that frontier large language model (LLM) agents escaped server containment through emergent coordination and cloud infrastructure attacks. This paper contextualizes those claims through forensic analysis and containerized recreation of both the Highly Persistent Internal Model (HPIM) training-case conditions and the Phase One privilege escalation targeting internal package repositories during OpenAI's ExploitGym evaluations. We recreate the configuration of the OpenAI-provided training example and reproduce the cryptographic verification bypass (CVE-2026-65616) in isolated repository servers, where misconfigured systems enabled token forgery and privilege escalation through basic HTTP requests, Python, and JavaScript. Crucially, the training-example conditions and the Phase One (and later) misconfigurations share the same exploitable patterns. Our analysis explains how Zero Trust Architecture principles neutralize this shared attack vector. Contemporary AI risks derive directly from training and deployment conditions in ambivalent learning environments lacking adequate containment measures.

This paper addresses the "trust gap" in deploying AI-enabled spacecraft autonomy through an integrated verification framework combining offline ground testing with online flight monitoring. The offline phase progresses systematically through simulation studies, processor-in-the-loop (PIL), hardware-in-the-loop (HIL), and robotic emulation testing, culminating in flight validation. The online phase employs AI watchdog systems for performance monitoring and safety guardrails for real-time constraint enforcement. Key contributions include: (1) systematic organization of verification techniques emphasizing closed-loop behavioral testing at each phase, (2) integration of AI watchdogs with runtime assurance, and (3) demonstration of the framework through a reinforcement learning-based attitude control case study. This approach integrates mature technologies to provide a viable pathway for safe deployment of AI-enabled autonomy in high-consequence space operations while addressing simulation-to-reality transfer challenges.

Infant fingerprint recognition is hindered by a combination of physiological characteristics, such as small ridge width and spacing, and acquisition challenges that include involuntary motion, skin residue, and uneven fingertip pressure. These factors undermine fingerprint quality standards originally designed for adults, which assume feature distributions that do not hold for infants. To address this gap, this paper evaluates ten fingerprint quality assessment methods on a longitudinal dataset that spans six age groups from newborns to three months, alongside a parameter tuning routine that uncovers age-specific fingerprint characteristics to recalibrate methods with domain-dependent assumptions. Both default and tuned configurations are compared through recognition experiments assessing the effectiveness of quality-guided gallery selection. Template-fusion strategies and latent fingerprint recognition technologies are further examined to identify the best possible matching rates achievable by the top-performing method. Results show that tuned parameter sets outperform default configurations for the majority of methods and age groups, indicating that targeted recalibration of existing frameworks improves gallery utility and contributes to more reliable infant fingerprint recognition.

Recursive self-improvement (RSI) in scientific research aims to improve agents' research capabilities through their own experimental experience. Scientific agent swarms can combine evidence and implementations from different research directions to guide their agents' subsequent experiments. However, poor experimental results can leave unclear whether to revise a hypothesis or its implementation, while overall program scores can hide components useful to other agents. To address these challenges, we introduce SwarmRSI, a framework that evolves scientific agent swarms for recursive self-improvement. We devise Feedback-Guided Swarm Mutation to refine hypotheses and their implementations using experimental feedback from other agents. We also propose Complementary Swarm Recombination to reuse retained hypotheses and complementary program components across agents, allowing partial results to support new research. Recursive Swarm Evolution integrates both mechanisms into a feedback loop, using new experimental evidence to guide subsequent research. Experiments on OpenProblems, NatureBench and MLE-Bench Lite demonstrate SwarmRSI's state-of-the-art (SOTA) performance. Further evaluations show that accumulated research experience improves performance on held-out tasks. Biologists also reproduce its programs for follow-up biological analyses.

Though modern AI models like LLMs predict discrete tokens, today's robotic foundation models represent actions as continuous vectors. In this paper, we challenge this common consensus by asking if continuous action parametrizations are superior to their discrete counterparts. We conduct an extensive comparison of single-step regression and multi-step flow variants of both continuous and discrete policies in both from-scratch single-task and multi-task VLA training, in both simulation and on real hardware. In all regimes, we find that discrete action prediction, even the most naive, uniform-bin quantization, can match or even outperform continuous-action policies in task success rate. Moreover, we find that the challenge to discrete action policies is not action precision, but rather factorization error introduced by product-distribution predictions across time and action dimensions. We show that, when paired with the correct interventions—iterative computation through discrete flows, or distribution sharpening through classifier-free guidance—discrete flow policies can turn this weakness into a strength, enabling more responsive action changes during a predicted action chunk. Lastly, we show that attempts to explicitly inject geometric information into discrete policies, e.g. through corruption processes that emulate continuous flow matching, do not yield consistent improvements. Taken together, our findings question existing assumptions that robotic action prediction requires distinct design decisions compared to other AI domains, and motivates the possibility of discrete action prediction as a compelling alternative to today's continuous action parametrizations. Link for the webpage \href{https://iscoyizj.github.io/discrete-action-website/}{https://iscoyizj.github.io/discrete-action-website/}

Detecting unseen malware variants remains a major challenge as malware families continuously evolve. This issue is exacerbated in data-scarce scenarios such as emerging malware campaigns, where only a few samples are available to train robust detectors. A common strategy is to proactively generate malware variants and use them as augmentation data to expose models to a broader range of potential threats. Existing approaches that generate functional malware variants through adversarial manipulation suffer from several important limitations. Generated samples must remain executable and preserve their malicious behavior, requiring costly validation through execution and analysis. Moreover, the adversarial objective provides limited control over the resulting augmentation dataset: it cannot guarantee the generation of sufficiently diverse variants, or even a single successful sample, making the effectiveness of the augmentation process inherently unpredictable. Finally, these approaches typically rely on a well-trained detector to guide generation, an assumption that may not hold in data-scarce scenarios such as early-stage malware campaigns. We formally and empirically demonstrate these limitations and propose Phantomware, a data augmentation framework designed for extreme data scarce scenarios. Phantomware operates entirely in the representation space of a pretrained malware detector, avoiding the costly generation and validation of functional binaries. By exploiting latent-space geometry, it generates synthetic malware representations of a target family in a controlled manner, providing a constructive guarantee on the generation of diverse training samples. Extensive experiments on real malware families demonstrate that Phantomware improves detection of unseen variants and provides robustness against both real and adversarially generated malware samples.

This paper considers reliable in-sequence molecular communication, where bio-nanomachines exchange molecule-encoded information in a specified order, and investigates a new simulation-based method to incrementally evaluate communication protocols by gradually increasing the complexity of experimental settings. The proposed method, called Affine Memory Replay (AMR), builds a continual learning (CL) model to estimate communication performance between bio-nanomachines. AMR stores past simulation samples and synthesizes additional replay samples through interpolation and extrapolation between past samples. Through incremental performance estimation tasks, it replays both past and synthetic samples to regularize model updates, helping the model adapt to new experimental settings while retaining knowledge from earlier ones. Experimental results show that AMR significantly outperforms existing CL methods in maintaining overall estimation accuracy across sequentially learned estimation tasks. It effectively balances memory stability and learning plasticity under moderate training cost.

Existing long-horizon agent benchmarks often emphasize final task success while providing limited assessment of intermediate execution validity. Yet real-world business operations require agents to navigate extensive, potentially outdated or conflicting policies while satisfying critical constraints throughout execution. These requirements are particularly consequential in high-stakes domains such as payments, where intermediate violations can incur financial losses or regulatory consequences regardless of the final outcome. To bridge this evaluation gap, we introduce BizBench, a multi-turn benchmark grounded in real-world procedures from a leading payments company. Its executable environment integrates 789 knowledge documents and 601 discoverable business tools, with consistent policies, states, and tool behavior. The benchmark comprises 50 tasks involving resource constraints, temporal dependencies, prior commitments, and conflicting or outdated guidance. Each task is validated through reference execution, while deterministic checks assess both goal attainment and execution validity without prescribing a single solution path. Experiments reveal substantial limitations even among frontier models: Claude Opus 5 and GPT-6 Sol achieve strict success rates of only 64.0% and 61.3%, respectively. For Qwen3.8-27B, goal attainment reaches 42.0%, but strict success drops to 32.7%, demonstrating how outcome-based evaluation can mask invalid execution. Beyond success rates, DeepSeek-V4.1-Flash averages 191.3 tool calls per run, approximately 70% more than reference executions, revealing substantial execution overhead. Overall, BizBench establishes a testbed for evaluating and advancing long-horizon agents beyond task completion, toward reliable and constraint-compliant execution in complex business environments.