Papers

The lasting gains in AI have come from methods that absorb more computation, and language models already scale computation within a call. Useful work, though, happens across calls: over hours or days, through tools and external systems, among many agents and the humans they work for, and increasingly through code the agents write themselves. Whether that work compounds is enabled by the harness around it, and the harness faces a constraint the model does not: of the three participants in such a computation, the human, the model and code, only the human's attention does not scale, so the computation can grow only if the human is not in every loop. L3M is a runtime built on that premise, written in Lean 4 so that one language implements the harness, states its invariants, and is the language agents extend it in. One computation spans humans, agents, models and code, and all of its state lives in Git: a rich state trajectory – rather than a chat conversation – is an object the runtime can resume, rewind, fork, compare and merge. Waiting, communication, delegation and undo are primitives, not patterns the model reasons through each time. When an agent works out a procedure, it can write it once as a sealed Lean program whose type is its permission request, and the harness offers it back as a tool – machinery for the model to use. To take the human out of the loop, L3M uses reversibility to decide where that is safe: a tool's review tier follows from how far its effects can be undone, so agents act freely where the runtime can recover. Because the machinery can change, the properties that make this safe are stated as kernel-checked theorems about the code that runs. On Terminal-Bench, an untuned L3M reaches about two thirds of its (unsafe) shell-equipped pass rate using only safe (not needing human review) tools.

Building on OpenAI's proof of isomorphism of free group factors, we give a self-contained proof of a corresponding result for $C^{*}$-algebras generated by free semicircular systems: the isomorphism class of a $C^{*}$-algebra generated by a semicircular $n$-tuple does not depend on $n\in\{2,3,\dots,+\infty\}$. Similarly, reduced free products of $n$ copies of the Jiang–Su algebra are isomorphic for different $n$ in this range. While reduced free group $C^{*}$-algebras of different ranks are not isomorphic (by a result of Pimsner–Voiculescu), we show that they have strong free absorption properties. At the heart of the proof is an automorphism approximation property for inclusions of $C^{*}$- and $W^{*}$-algebras; a relative version of that property was previously used to derive free complementation results.

Repeated event forecasting requires extracting predictive signals from noisy evidence and evaluating their value across successive decisions. We introduce JockeyBench, a benchmark and evaluation protocol that studies this problem through historical Japanese horse racing. The protocol combines time-gated evidence, priced tools, paired frozen-program controls, and two workflows: executing an LLM-written predictor or acquiring evidence for a fresh per-race LLM forecast. Evaluation separates forecast quality, benchmark-specific net utility, and retrospective betting returns and risk under fixed decision rules. Across 28 model–scaffold configurations on 324 races, paired tests identify seven revision gains and two declines after multiple-testing correction. Yet every configuration has negative mean profit under capped-Kelly settlement. In a four-model comparison, the reusable-program workflow has higher mean betting profit and lower research fees, although both workflows lose on average. Further diagnostics reveal that mean probability bias reverses on selected bets and that profitable individual runs can depend on concentrated returns. These findings motivate evaluating predictive competence, adaptation, resource use, and decision value separately when assessing agents for repeated event forecasting.

We consider the problem of estimating the background intensity of a Hawkes process when the triggering mechanism is known, treating the background as an unknown finite positive measure with a regular density. We propose a likelihood-based variational approach based on the Wasserstein-Fisher-Rao geometry. In this setting, the relevant first variations are naturally distribution-valued and therefore fall outside the classical gradient-flow formulation. We develop a distributional framework for the resulting dynamics and give them a finite-dimensional meaning through projections onto statistical manifolds. We identify the true background as the unique zero of a limiting generalized gradient and, under suitable stability, compactness, integrability and approximation conditions, establish a conditional consistency theorem for measurable estimators satisfying the empirical projected equations on events whose probabilities tend to one. Finally, we specialize the construction to Gaussian-mixture manifolds and obtain an explicit finite-dimensional scheme.

Social psychology has long described how people can be persuaded to interpret a harmful act as moral. Recent research suggests that Large Language Models (LLMs) echo aspects of human psychology, from cognitive biases to social conformity. This raises a question with direct safety implications: can LLMs be moralized to bypass safety guardrails? We answer this question with a multi-agent framework grounded in Moral Foundations Theory (MFT). Given a harmful request that is initially refused, our agents diagnose which moral foundation (MF) the prompt is violating, identify a justifiable foundation under which the same request appears virtuous, rewrite the prompt accordingly, and reinforce it with generated precedents of comparable moral reframing. We evaluate our framework on harmful prompts from three well-known jailbreak datasets across five open-source and proprietary models. Our analysis reveals that models are highly susceptible to moralization, exhibiting high single-trial jailbreak rates with success increasing further under repeated attempts. We also find that safety benchmarks are imbalanced in the MFs they violate, concentrating on foundations toward which models already exhibit biases.

We construct a seven-cycle example for a periodically forced scalar equation with a continuous three-zone piecewise-linear nonlinearity. Symbolic integration and linear elimination reduce the four-flight matching conditions to two equations in the residence times. The competing exterior visits produce an additional turn of the critical response and lead to the rational parameters $(A,B,\mu)=(9,3/2,181363/100000)$. At these parameters, finite Taylor estimates give three sign changes of the return displacement. Scalar order, analyticity along transverse itineraries, and half-period reflection then establish at least seven distinct isolated periodic orbits on the phase cylinder.

In this work, we propose a curve to be a minimum of length among rectifiable, continuous curves of unit width in three-dimensional Euclidean space. We show that a previously conjectured minimizer of that problem is not of unit width, give the numerical derivation of a candidate family of unit width curves, prove that it admits a strict local minimum within that candidate family and that this minimum is of unit width and length $3.905124162243593\ldots$. The curve consists of eight straight segments, two cone geodesics, and three circular helices. Its parameters are specified by a uniquely isolated solution of the joining and stationarity equations. To establish unit width in every direction, we combine analytic estimates with interval arithmetic on a finite covering of the remaining directions. We conjecture that the curve is globally minimizing. We supplement this work with the programs used to identify the candidates, replicate the computer-assisted parts of our proof, as well as a formalization of our proof in Lean.

Greibach and Ginsburg (1969) studied abstract families of languages or AFLs. A full AFL is a class of languages closed under the following operations: homomorphism, inverse homomorphism, intersection with regular languages, union, concatenation, and Kleene star. Greibach (1970) used a restricted substitution, called nested iterated substitution, to introduce the notion of super AFLs, which are full AFLs that are further closed under nested iterated substitution. The smallest super AFL was shown to be the class of all context-free languages. From this fact, we focus on Greibach's limited, univariable application of nested iterated substitution and further generalize it into k-variate. This introduces the notion of nested iterated $k$-variate substitution whose definition does not rely on "machines". Recently, a "two-tape" version of Hibbard's (1967) rewriting systems, which were later reformulated by Pighizzini and Pisoni (2014) as k-limited automata with "single" input/work tapes, has been studied under the name of depth-k storage automata. This work unearths a few characteristics of the languages recognized by those automata under a restriction known as frozen blank sensitivity in terms of full AFLs and nested iterated k-variate substitution. This will promote a further abstract treatment of those languages.

As large language models (LLMs) are increasingly used to generate text, their misuse has become a growing societal problem, with implications for academic misconduct, model distillation, and cyberattacks. For these reasons, Model Authorship Attribution (MAA), which identifies the source model of a given text, is important for promoting the responsible use of LLMs and improving AI safety. In practice, attribution systems must remain reliable beyond the distributions observed during training. They may encounter texts from unseen domains and writing settings, as well as generations from future versions of model families that were not available when the system was trained. However, existing MAA methods often exhibit substantial performance degradation under such distribution shifts. To address this challenge, we propose Authorship Tracing via Likelihood-based Attribution of Sources (ATLAS), which formulates MAA through source-conditioned text modeling rather than direct source prediction. ATLAS learns how text from each source family is generated and attributes a given text by comparing its likelihood under each learned source condition. We aggregate this likelihood evidence at the token level with clipping so that no single token dominates the decision. We further introduce MAABench, a broad MAA benchmark for evaluating practical OOD settings such as domain and version. Our experiments show that ATLAS consistently outperforms existing baselines across MAABench and OpenTuringBench. In particular, when both the generation domain and the model version change on MAABench, ATLAS achieves 74.27% accuracy, 9.00pp higher than the strongest baseline. Beyond MAA, ATLAS can also help identify distillation source models by measuring which source model a target LLM most closely resembles. Our extended experiments show that ATLAS recovers the known teacher of every distilled model we evaluate.

Magnetic proximity effects in transition metal dichalcogenide (TMD)/2D-magnet heterostructures are often described in terms of the magnetization of the interfacial magnetic layer. Here, we show that this picture is sufficient to describe the neutral-exciton valley splitting of the TMD, but incomplete for its hole-doped electronic structure. We study a gate-tunable van der Waals heterostructure composed of monolayer WSe$_2$ in direct contact with trilayer CrI$_3$. By combining spatially and helicity-resolved reflection-contrast spectroscopy with magnetic circular dichroism, we identify two distinct magnetic domains in the trilayer CrI$_3$: one with purely antiferromagnetic interlayer coupling and one with a mixed antiferromagnetic (AFM)/ferromagnetic~(FM) coupling. In both domains, the sign of the neutral-exciton valley splitting of WSe$_2$ switches with the magnetization reversal of the adjacent CrI$_3$ layer. In contrast, helicity-resolved positive-trion spectroscopy of WSe$_2$ reveals a qualitatively different response. While the AFM domain shows a strongly helicity-selective trion resonance, the mixed AFM/FM domain exhibits trion resonances in both helicities. This behavior disappears when all CrI$_3$ layers are aligned in a parallel magnetization direction. Our results demonstrate that charged excitons are sensitive to the full layer-resolved magnetic sequence of the adjacent 2D-magnet, revealing magnetic proximity beyond the interface.

Large language model (LLM) agents can reuse experience from prior task executions to improve performance on subsequent tasks. Although existing methods enable cross-agent experience sharing, differences in agents' execution histories and task requirements make shared experiences difficult to reuse effectively. We propose XferAgent, a framework for cross-agent experience transfer that organizes reusable experience in local and shared experience graphs. XferAgent reconciles compatible contributions from heterogeneous agents while avoiding unreliable merges. For each new task, it selects shared experiences and integrates them with the target agent's local experience to support task execution. Experiments across diverse task domains show improvements over representative memory-based methods and effective experience transfer between heterogeneous agent frameworks. Cumulative task success also increases over successive task passes as agents update and reuse their experiences.

We study a PDE-constrained variational problem for the optimal spatial distribution of a catalytic material reacting with a substrate in a biological or materials system. With a given catalyst concentration, the substrate concentration is determined by a reaction-diffusion equation. The reaction functional for the catalyst concentrations is defined to be the total reaction output, which is the product of catalyst and substrate concentrations. Both the linear and Michaelis-Menten reaction kinetics are considered. We construct reaction-maximizing sequences through the localization of catalysts using barrier functions and show the non-existence of a global maximizer. We also derive the first and second variations of the reaction functional. By constructing a competing catalyst concentration, we show the non-existence of a local maximizer. Based on our analysis, we propose a regularized variational model and prove the existence of its global maximizer.

State-of-the-art visuomotor policies are increasingly capable of executing complex tasks, yet can fail during deployment in ways they cannot predict or recover from on their own. Improving their performance and robustness typically requires collecting more data, retraining, or modifying the policy architecture, all of which are time- and resource-intensive in practice. We instead propose an offline-trained monitoring-and-intervention framework that improves a policy during deployment using only black-box query access. From a limited set of labeled on-policy rollouts, we learn a policy-conditioned liveness value that measures the nominal policy's progress toward task completion; we additionally train a compact latent world model using these rollouts together with play data. At deployment, the liveness value continuously monitors execution. When it detects a loss of progress, a short-horizon planner searches the world model for corrective actions using the same value as its planning objective, after which control returns to the nominal policy. We evaluate our method on the LIBERO-Goal task suite and on a Franka Panda manipulator, using the vision-language-action model $\pi_{0.5}$ and a diffusion policy as nominal policies, and demonstrate significant improvements in task success rates, while supporting real-time control. Additional results and videos are available on the project website (https://policy-improvement-wm.pages.dev/).

Graph Collaborative Filtering models have achieved remarkable success in recommendation systems. However, they inherently suffer from a catastrophic representation collapse in strict cold-start scenarios due to the topological isolation of unobserved items. While integrating multi-modal semantic features offers a potential remedy, directly mapping isolated semantic attributes into a globally topology-driven collaborative space frequently precipitates representation degradation and optimization conflicts. To bridge this fundamental structural divide, we propose DualGCN, an innovative Dual-Branch Graph Knowledge Distillation Framework. Operating under a decoupled two-stage training paradigm, DualGCN first constructs a robust collaborative teacher network where item representations are anchored by a topology-smoothed semantic prior, effectively reshaping the latent manifold. Subsequently, a semantic student branch employs a residual mapping mechanism and a magnitude-agnostic cosine distillation objective to achieve high-fidelity cross-modal alignment without disrupting established collaborative signals. Furthermore, we introduce a dynamic norm reconstruction mechanism to rigorously calibrate the physical scales of novel and warm items during global retrieval. Extensive experiments across three real-world benchmark datasets demonstrate that DualGCN significantly outperforms state-of-the-art baselines in strict cold-start scenarios while comprehensively preserving recommendation accuracy for highly active items.

Using fully coupled two-way simulations combining the vortex filament model with the Navier–Stokes equations, we discover that quantized vortices in superfluid $^4$He are systematically rolled up into coherent bundles within normal-fluid vortex tubes. Although co-rotating quantized vortices do not aggregate spontaneously, mutual friction with the normal fluid actively drives their bundling, directly inheriting the spatial hierarchy of classical turbulence. Mutual friction excites high-wavenumber normal-fluid vortex tubes adjacent to quantized vortices, while the superfluid forms a nested structure of thick low-wavenumber outer tubes and fine high-wavenumber inner tubes around the vortex lines. A distinct time lag between superfluid kinetic energy and dissipation fuels the energy cascade, giving rise to Kolmogorov's $k^{-5/3}$ spectrum. Crucially, bundle lifetimes are governed by the ratio of the local coupling time scale $t_\mathrm{cl}$ to the Kolmogorov dissipation time scale $\tau_\eta$ ($t_\mathrm{cl}/\tau_\eta < 20$), leading to enhanced thermal stability at higher temperatures. These findings establish a unified mechanism for two-fluid multiscale vortex organization.

Temporal graph data are recorded in different representations, but whether these differences also determine the scope of learned knowledge reuse remains unclear. This study investigates whether predictive structure learned from relational histories can be reused between continuous-time dynamic graphs (CTDGs) and temporal knowledge graphs (TKGs). Bidirectional reuse emerges while retaining distinct input and prediction interfaces and independent vocabularies, and sources with the same representation are not always more useful. Controlled reorganizations of source histories show that the organization contributing to reuse varies across source-target conditions. We show that learned aspects of Relational Dynamics, the predictive structure present in relational histories, can be reused across conventional representation boundaries. Independently pretrained TKG and CTDG learners can also be fused at the score level to improve prediction on a third dataset without entity or relation-semantic alignment. Two proofs of concept in recommendation and security detection illustrate that external temporal sources without direct alignment can also become useful pretraining resources.

We perform a comprehensive analysis of a radion/dilaton portal freeze-in dark matter in models of stabilized warped extra-dimensions. We show that the relic abundance of a generic dark matter particle (scalars, vectors, or fermions) is obtained via the conformal coupling of the radial mode in a consistent model stabilized warped extra- dimensions; our calculations show that in scenarios where the reheating temperature is much smaller than the radion mass, keV-TeV scale dark matter particles can saturate the relic density. We perform a comprehensive analysis of a radion/dilaton portal freeze-in dark matter in models of stabilized warped extra-dimensions. We show that the relic abundance of a generic dark matter particle (scalars, vectors, or fermions) is obtained via the conformal coupling of the radial mode in a consistent model stabilized warped extra- dimensions; our calculations show that in scenarios where the reheating temperature is much smaller than the radion mass, keV-TeV scale dark matter particles can saturate the relic density. Furthermore, we show that the model is subject to Lyman-alpha constraints in the keV-GeV range, providing a unique, model-independent probe of this sector. Our calculations extend previous attempts to calculate the freeze-in relic abundance in extra-dimensional portals by providing a consistent theoretical calculation of matrix elements, clearly laying out the underlying motivation and assumptions and providing a comprehensive account of the cosmological constraints on the model.

A Zumkeller number is a positive integer whose set of positive divisors can be partitioned into two disjoint subsets of equal sums. In this paper, we investigate the additive properties of these numbers in conjunction with prime numbers. Specifically, we study the natural density of the exceptional set \(\mathcal{E}\) consisting of all positive integers not expressible as the sum of a prime and a Zumkeller number. Let $\alpha(x) = \#\{n \le x : n \in \mathcal{E}\}$ denote its counting function. We establish a quantitative upper bound for $\alpha(x)$, demonstrating that $\alpha(x) \ll x(\log \log x)^{-A}$, where $A = \log_2\left(\frac{2\pi^2}{2\pi^2-3}\right) \approx 0.2378$, thereby proving that the natural density of this exceptional set is $0$.

Differential Privacy (DP) provides a rigorous framework for quantifying privacy guarantees. While most methods achieve DP by injecting calibrated random noise, the intrinsic randomness of certain procedures such as sampling can yield DP guarantees “for free”. In this work, we develop a unified Rényi divergence framework to characterize the inherent DP guarantees of sampling from a Bayesian posterior distribution. We show that the privacy loss incurred by One-Posterior-Sampling (OPS) is governed by posterior exponential moments of the single-record log-likelihood ratio, and establish explicit DP guarantees under uniformly bounded, sub-Gaussian, or sub-exponential tail regimes. We apply the framework to representative Bayesian models, including categorical likelihoods with arbitrary priors, Gaussian-Gaussian conjugate pair, generalized linear models from the exponential family and linear regression with Gaussian likelihood, both with Gaussian priors on the regression coefficients. Our results elucidate sample size, model structure, neighboring relations in DP (bounded or unbounded), and how prior if applicable, jointly determine the inherent DP guarantees of OPS. Furthermore, our results recover the existing DP guarantees for OPS as special cases – often with tighter privacy loss bounds – and substantially broaden the class of Bayesian models for which the inherent DP guarantees of OPS can be rigorously characterized.

On-policy distillation (OPD) transfers knowledge from a teacher model to a student along student-generated trajectories, typically by minimizing the reverse Kullback-Leibler (RKL) divergence. However, RKL can provide weak corrective signals for teacher-supported tokens to which the student assigns very low probability. This raises two questions: which positions are most consequential for the behavioral gap between the student and teacher, and what additional supervision is needed at these positions? Through cross-sampling, we find that divergence-based routing more effectively pinpoints positions contributing to the teacher-student performance gap than entropy-based routing. Further analysis reveals teacher-mode under-coverage at high Jensen-Shannon (JS) divergence positions, where standard OPD increases the large teacher-student probability gaps. Motivated by these findings, we propose Divergence-Routed On-Policy Distillation (DrOPD). DrOPD retains RKL supervision at all positions and selectively adds forward Kullback-Leibler (FKL) divergence supervision at positions with high JS divergence. Experiments across six math benchmarks show that DrOPD improves over standard OPD and entropy-based routing. DrOPD also reduces severe under-coverage cases by 37.9 percent relative to OPD, supporting targeted FKL as an effective complement to OPD.

Real-world engineering drawings are often large-format and information-dense. Answering practical engineering questions therefore requires both fine-grained understanding of engineering information and complex reasoning over objects, relations, and evidence distributed across different regions or views. Existing benchmarks cover a range of engineering-drawing understanding tasks involving local regions, individual views, specific elements, or more constrained visual settings. However, systematic evaluation of understanding over realistic, large-format engineering drawings remains limited. To address this gap, we introduce LED-BENCH, a multidisciplinary benchmark designed to evaluate fine-grained understanding and complex reasoning over large-format engineering drawings, spanning the mechanical, architectural, and transportation domains. Extensive evaluation of 11 representative multimodal large language models, including 8 proprietary models and 3 open-source models, reveals substantial remaining difficulty: GPT-5.6 Sol achieves 52.65% average accuracy, performance varies across engineering domains, and all evaluated models perform worse on questions requiring broader evidence integration. These results highlight fine-grained understanding, evidence integration, and complex reasoning over large-format engineering drawings as persistent challenges for current multimodal models.

Quantifying the diversity of a set of objects is a fundamental problem in machine learning and statistics. Mironov and Prokhorenkova (2025) recently showed that three axioms-monotonicity, uniqueness, and continuity-are crucial for a reliable diversity measure and that none of the standard metrics in the literature satisfy all three. They also constructed two measures that do satisfy the axioms, but both are NP-hard to compute, leaving open the question of whether efficient axiomatic diversity measures can exist. We study a Euclidean-restricted version of this question. We introduce Geometric Kernel Diversity (GKD), which associates a radial kernel solid with each point and measures the volume of their union. We show that canonical GKD satisfies all three Mironov-Prokhorenkova axioms on Euclidean-realizable inputs. For canonical GKD, we give exact polynomial-time algorithms in dimensions $d \le 3$. In arbitrary dimensions, we give a randomized algorithm that returns a $(1 \pm \varepsilon)$-approximation with probability at least $1-\delta$ in time $O \left(\frac{nd}{\varepsilon^2}\log\frac{2}{\delta}\right)$ for $n$ points. The axiomatic guarantee applies to exact canonical GKD; the randomized algorithm approximates this deterministic quantity. We further show that the two diversity measures of Mironov and Prokhorenkova (2025) remain NP-hard on Euclidean-realizable inputs and, unless $\mathrm{NP} = \mathrm{RP}$, admit no FPRAS even on this restricted domain.

This paper reconstructs the Michelson interference phenomenon based solely on the binary (on-off) statistics of a single-photon counter and macroscopic path-difference data, without the a priori assumptions of complex wave superposition or prior spectral information. In contrast to the conventional approach that assumes light as a continuous complex wave function and considers interference between them, this study focuses on the algebraic properties inherent in discrete detection statistics. We verify that the discrete response data collected across various path differences form a positive semi-definite matrix, and apply Herglotz's theorem to this result. Consequently, we show that the harmonic form of the interference fringe is not a consequence of adopting a wave model as an axiom, but rather it is uniquely decomposed into a cosine average of wavenumbers dictated by testable algebraic constraints satisfied by the statistical data. Through this approach, the logical basis of the interference phenomenon can be described by the characteristics of the binary data itself, perfectly aligning with the modern optical experimental environment that relies entirely on discrete event statistics.

Latent reasoning replaces natural-language chains of thought with compact continuous "soft-token" representations, reducing the number of intermediate tokens that must be decoded autoregressively. We use a small external encoder to induce latent reasoning in a frozen backbone model. The encoder iteratively maps the backbone's hidden states to new latent blocks, constructing a latent reasoning prefix, after which the frozen model writes its remaining reasoning and its answer as text. Our method, INTACT, trains this encoder through distillation to reproduce the effect of written reasoning steps while keeping the model's behavior close to that of the frozen backbone, and then reinforces it on answer correctness. Across four frozen backbones from 1B to 8B, INTACT improves both GSM8K and average out-of-domain accuracy, including GSM8K gains from 75.8% to 79.3% on Llama-3.2-3B and from 77.6% to 79.7% on Qwen2.5-3B. It also reduces end-to-end latency per problem by 19% on Llama-3.2-3B and outperforms the evaluated latent-reasoning baselines at the 3B and 8B scales. Beyond the mathematics setting, the same recipe trained on a multi-domain corpus improves the frozen backbone on twelve evaluation benchmarks spanning mathematics, science, and commonsense, without task-specific adaptation.

We present self-consistent $N$-body/hydrodynamical simulations of isolated, Milky Way-like galaxies with stellar and gaseous disks. By varying the initial gas fraction, we find that gas delays bar formation and, in the most gas-rich model, suppresses it by exciting higher-order spiral modes and promoting early central mass growth. Once a bar forms, continued gas inflow and gravitational torques weaken the bar and inhibit its secular slowdown, keeping bars in the gas-bearing models shorter and predominantly in the fast-bar regime, while the gas-free bar evolves into the slow-bar regime. Stars and gas respond differently to the same barred potential. While the mean stellar velocity field is well described by a smooth bisymmetric pattern, the cold and warm gas follow elongated, $x_1$-like streamlines with sharp shock-induced velocity transitions, producing sawtooth-like radial and archlike tangential velocity profiles. Despite these strong noncircular motions, their azimuthally averaged tangential velocities closely follow the circular velocity over most of the disk. By contrast, the hot gas is dominated by stellar feedback and thermal pressure and is largely decoupled from the ordered bar-driven flow. To describe these bar-driven gas flows, we introduce the barflow model, a phenomenological description of non-sinusoidal gas streaming motions with a radially varying shock-front phase. Mock line-of-sight velocity-fitting experiments show that the barflow model recovers the intrinsic rotation curve, bar-driven velocity perturbations, and radial gas transport more accurately than the conventional bisymmetric model.

Autonomous drones are increasingly deployed in safety-critical tasks, yet their autopilot software can still contain subtle logic flaws that lead to unsafe behaviors. These flaws are difficult to discover with existing techniques because they often depend on semantic relationships across modules, commands, and runtime state, and rarely manifest as crashes or detectable failures. With recent advances in LLMs, we observe that, when equipped with essential domain knowledge and carefully crafted context, LLMs demonstrate strong capabilities in understanding the high-level semantics of autopilot systems, making them suitable for logic flaw discovery. However, LLMs are limited by their context and can be biased to the vulnerabilities frequently occurs in training dataset, while logic flaws often require higher-level semantic bug oracles to identify their existence. To bridge the gap, we present Impossible Mission Finder (IMF), a multi-agent system combining static analysis, dynamic execution and log analysis, that supports fully automated identify, test, and validate logic flaws in autonomous vehicles. IMF primarily contains three LLM agents, it first uses the auditor agent to propose logic flaw hypothesis based on carefully-crafted program slicing, then it launches tester agents to write and execute test cases for each candidate, and finally it spawns the log analyzer agent to validate the execution log against the flaw hypothesis. This enables IMF dynamically validate bug oracles in a case-by-case manner, making it suitable for discovering logic flaws that do not have a unified pattern. We evaluated IMF on two widely used open-source autopilot systems, PX4 and Ardupilot, leading to 105 previously unknown logic flaws, that can cause unsafe behaviors including mission interruption failures, invalid actions, entering into restricted fly zones etc.

The rapid growth of AI data centers introduces large, workload-dependent loads whose protection and reconnection behavior can affect power-system transient stability. Calibrating dynamic models of such loads is challenging because treating all associated elements as freely estimable may lead to weakly identified or physically confounded parameterizations. This paper proposes an identifiability-guided calibration framework for a composite load model with separate IT, cooling, and auxiliary branches. A nuisance-projected Fisher-information criterion characterizes the local identifiability of the cooling-motor composition, while a workload non-equivalence result shows that voltage-only recalibration cannot reproduce workload-dependent responses. Guided by these results, a grey-box estimator trained with branch-level supervision learns the workload-to-power relation and the cooling composition, while retaining a prescribed UPS protection law and an externally anchored reconnect-ramp duration. The learned workload relation is then distilled into a closed form for transient-stability simulation. Evaluation using 300,000 MIT Supercloud GPU power-utilization samples and 12,600 IEEE 39-bus fault scenarios yields an aggregate active-power $R^2$ of 0.948, compared with 0.856 for an LSTM baseline. Removing branch-level supervision preserves relatively high aggregate accuracy but substantially degrades cooling-component recovery. Under unseen deep voltage sags, the proposed model achieves a transient $R^2$ of 0.750, compared with 0.033 for the LSTM. Closed-loop simulations further show that cooling composition and protection behavior can affect retained-load states and voltage responses near protection thresholds.

A spatial answer depends on the reference frame specified in language and its supporting visual evidence. Item accuracy alone obscures whether correctness persists when either changes. We introduce Frame Evidence, a benchmark of 320 matched triplets and 960 Chinese multiple-choice questions from captured and rendered scenes. One family compares three grounded reference frames over fixed images. The other pairs image deletion annotated as insufficient with answer-preserving addition. Joint metrics require correct answers across all three frames and paired evidence variants. Across 20 configurations on 288 held-out questions, GPT three-shot reaches 50.00% item accuracy, but the best three-frame group accuracy is 5/48 (10.42%), reached by GPT two-shot. Few-shot prompting and task-specific adaptation improve item accuracy without consistently improving joint correctness. Input controls show where images support matched success, while repeated executions reveal answer instability beneath similar scores.

Safety alignment of text-to-image diffusion models aims to suppress harmful content without compromising useful generation capabilities. However, preferences over completed images compress multidimensional generation factors into a coarse signal, potentially driving changes beyond those required for safety. To address this, we propose Safety Transition Preference Optimization (SaTPO), which enforces step-wise alignment during denoising. SaTPO leverages the insight that an unsafe prompt and its safe rewrite naturally form a tightly coupled preference pair when conditioned on a shared parent state and noise realization. By anchoring optimization on identical parent states, SaTPO eliminates historical trajectory divergence, confining safety adaptation to an incremental alignment while preserving task-relevant semantics. Multi-scale drift constraints and benign replay further limit behavioral deviation and preserve ordinary generation. We additionally construct CoProPV, a refined dataset ensuring visually verified contrasts. Experiments on Stable Diffusion v1.5 and SDXL demonstrate improved safety-quality trade-offs. Compared with AlignGuard, PerspectiveVision unsafe-detection rates on Inappropriate Image Prompts decrease from 12.70% to 6.00% and from 15.04% to 9.30%, respectively, while FID improves from 32.28 to 25.95 and from 36.50 to 27.51. Warning: this paper includes potentially offensive content.

Parikh's famous theorem relates to the semi-linearity of context-free languages. We seek for expanding the scope of the semi-linear property from context-free languages to languages beyond them. Of those languages, this work focuses on the languages that are particularly recognized by one-way nondeterministic depth-k storage automata under a natural restriction of frozen blank sensitivity. Those languages (which are called $k$-storable languages) have been already proven to satisfy a natural generalization of the well-known pumping lemmas for regular and context-free languages. The semi-linearity of k-storable languages generalizes Parikh's theorem on the semi-linearity of context-free languages. As for bounded languages, Ginsburg and Spanier's well-known theorem characterizes context-free languages in terms of stratified semi-linearity. This notion is further generalized and is used to characterize the k-storable languages as well. Such a generalization is interesting in its own right. This helps us take an abstract treatment of k-storable languages without relying on "machines". Our results strengthen the belief that those languages are "natural" generalizations of context-free languages.