Papers

Mufwene's ecological model locates language evolution in competition among variants contributed by individual idiolects and in speakers' selection from linguistic material made available through interaction. Large language models (LLMs) complicate this architecture without requiring the locus of selection to move away from human speakers. This article argues that LLMs are best treated as distributional mediators: they aggregate language produced across human populations, transform its distribution through training and post-training, and redistribute model-specific outputs at scale. I call the resulting ecological process algorithmic reweighting of the speaker-accessible distribution: model mediation can alter the relative frequencies with which competing variants reach human selectors. Emerging evidence on model-specific linguistic profiles and lexical uptake is consistent with parts of this pathway, but does not establish inevitable convergence. Human social evaluation remains decisive: model-associated forms may diffuse and become conventionalized, become socially recognizable as 'AI-like' and subsequently avoided, or fail to diffuse in the first place. The proposal extends Mufwene's feature-pool ecology one step upstream of speaker selection and yields testable predictions about uptake, model-version effects, convergence, and social reversal.

Large language models have significantly enriched the toolkit for computational humor research, particularly in the automated generation of jokes and puns. A key innovation, contextual embedding vectors, offers new opportunities to revisit and refine earlier hypotheses. Notably, Petrovic and Matthews (2013) proposed a joke generation model based on the scheme "I like my X like I like my Y, Z" (e.g. "I like my ice like I like my dreams, crushed"). They suggested that joke hilarity increases with: a) frequent association of Z with X and Y, b) rarity of Z, c) ambiguity of Z, and d) meaning distance between X and Y. Building on this, Winters et al. (2019) proposed a set of metrics, based on Google Ngrams and Word2Vector. In this work, three out of their five metrics are revisited with word embeddings: obviousness, compatibility, and comparison. Another measure, symmetry, defined as closeness of Z to both X and Y, is introduced here for the first time. Two models were used to collect the embedding vectors (OpenAI text-embedding-3-small and MiniLM all-MiniLM-L6-v2) on three datasets: JokeJudger, Expunations, and rJokes. The last two datasets, Expunations, and rJokes, were expanded by adding paired sentences that captured the ambiguous expression at the core of each joke in its two different meanings. Results revealed that models trained on the proposed metrics performed poorly in predicting humor ratings: on JokeJudger, the best model achieved 57.1% accuracy, below the 61.5% baseline, while performance on Expunations and rJokes was even lower. Nevertheless, the symmetry metric seems consistently associated with higher-rated jokes, suggesting it may capture a necessary -though not sufficient- property of humor.

Successful Human-Robot Teaching assumes alignment between robot processing needs and human teaching intent. To better understand this alignment, this work seeks to uncover the underlying logic that humans intuitively apply when teaching. Through an exploratory, bottom-up study with N=34, participants observing two distinct robot Reinforcement Learning (RL) scenarios, we analyze 204 intuitive teaching responses across early, middle, and late learning phases. Results reveal that teaching decisions consist of a nuanced, interconnected network of Triggers (situational catalysts), Objectives (subjective teaching targets), Signals (communicative acts), and Strategies (high-level governance) in which teachers spontaneously adopt diverse roles, acting as coaches, engineers, or designers and prioritize different objectives. Based on these results, we introduce the TOSS Framework, which conceptualizes Human-Robot teaching as a procedural loop between robot behavior and human teaching actions, in which human teaching decisions are modeled as Trigger-Signal responses modulated by teaching Objectives and Strategies. It provides future research with an openly accessible dataset and a theoretical foundation for a) understanding teaching decisions and b) simulating realistic oracles as well as c) designing human-centered teaching settings and novel robot learning algorithms that go beyond the constraints of current robot learning settings.

Courcelle's Theorem states that on graphs $G$ of tree-width at most $k$ with a given tree-decomposition of size $t(G)$, graph properties $\mathcal{P}$ definable in Monadic Second Order Logic can be checked in linear time in the size of $t(G)$. Inspired by L. Lovász' work using connection matrices instead of logic, we give a generalized version of Courcelle's theorem which replaces the definability hypothesis by a purely combinatorial hypothesis using a generalization of connection matrices. This paper clarifies the role of logic in such theorems and displays their purely combinatorial assumption.

Adverse Pregnancy Outcomes (APOs) such as preterm birth and gestational diabetes can have long-term consequences for both the mother and child, yet an understanding of their causes remains elusive. Causal discovery in this domain is especially challenging due to a paucity of data and incomplete domain knowledge. As a result, pure data-driven methods fail, and Large Language Model (LLM) outputs remain inconsistent or contradictory. We introduce a neurosymbolic framework for generating plausible causal hypotheses that iteratively combines the broad prior knowledge of LLMs with empirical scoring on data. Our method treats the LLM as an adaptive proposal distribution, generating hypotheses that are scored against empirical data; the resulting high-scoring graphs are then used to update the LLM's context, steering subsequent generations toward more promising regions of the hypothesis space. We evaluate our approach on a real-world clinical dataset for modeling APOs and their risk factors, comparing our results against an expert-constructed causal graph. Our method recovers all expert-validated edges and identifies additional plausible causal relations not previously listed by experts, potentially providing new insights for targeted interventions.

This technical report presents AudioWorldSim, an open-source platform designed to generate realistic binaural audio datasets and advance research in audio-based machine learning, particularly world models. Built as a custom extension of Meta's SoundSpaces 2.0 platform, AudioWorldSim leverages their comprehensive acoustics framework, but focuses on the automatic rollout of random agent navigations, as well as implements crucial fixes to how continuous sound is composed. AudioWorldSim is made publicly available to the research community at https://github.com/Luizerko/AudioWorldSim to facilitate reproducibility.

Large language models (LLMs) are increasingly used in research workflows and software development pipelines, yet their output remains sensitive to input prompt variations. This paper presents $\unicode{x00AB}$PromptResponse$\unicode{x00BB}$, a controlled study examining how formatting and LLM-based tuning of coding task prompts affect the resulting code's performance, efficiency, and stability. Using five semantically identical yet syntactically distinct variants of the HumanEval dataset$\unicode{x2014}$baseline, JSON, Markdown, YAML, and an LLM-tuned version$\unicode{x2014}$we had GPT-4o solve its coding problems over 8200$\unicode{x00A0}$executions. Our results show that consistent formatting$\unicode{x2014}$especially JSON$\unicode{x2014}$improves generation efficiency and syntactic stability, with minor gains in task performance. Conversely, the LLM-tuned prompts resulted in significantly degraded task performance without significant improvements in any other dimension. These findings suggest that low-effort reformatting alone can yield measurable improvements, while tuning must account for model alignment. We conclude our work with providing a set of practical recommendations informed by our results as well as releasing our dataset variants and evaluation pipeline for future work.

Inferring continuous system evolution from sparse temporal snapshots is a key challenge in generative modeling and single-cell omics. While Optimal Transport (OT) is popular, existing frameworks are largely restricted to first-order dynamics, assuming memoryless velocity fields. This limits expressiveness, as first-order systems fail to account for regulatory momentum and time-delayed responses inherent in processes like cell differentiation. Here, we introduce TracingFlow, a simulation-free Flow Matching framework generalizing to second-order dynamics. By using neural networks to regress the acceleration field, TracingFlow provides an exact, efficient solution to the Dynamical Optimal Acceleration Transport (DOAT) problem. Unlike first-order methods yielding over-smoothed trajectories, our second-order formulation captures high-curvature transitions and nonlinear evolutions by learning the underlying force fields. Evaluated on complex synthetic and large-scale scRNA-seq datasets, TracingFlow achieves superior accuracy in distributional reconstruction and trajectory faithfulness. Moreover, by integrating lineage tracing priors, it recovers dynamical structures that are both mathematically optimal and biologically plausible.

LLM-assisted agents can synthesize system evidence, propose configuration changes, and automate diagnostic tasks, but their flexibility makes an imprecise action or an intrusive collector an operational risk. We present Spike-Killer, a human-approved workflow for diagnosing frame-time complaints on one real Windows workstation. The workflow treats each action as an evidence-gated transaction: it records the exact target state, classifies risk, preserves a snapshot, verifies a postcondition, and retains failed measurements as first-class evidence. This experience paper reports a completed same-day study with Counter-Strike 2 as a demanding target application. The evidence bundle contains preserved state snapshots, exploratory microbenchmarks, a ten-run same-state repeatability probe, live telemetry, a repaired over-broad registry action, incompatible presentation-capture attempts, an invalid local replay, and a system-level tracing replacement. Windows Performance Recorder produced two CS2 local-Bot GPU traces of 90.69 and 85.85 seconds; both were attributed to cs2.exe, exposed DxgKrnl Present metadata, and had zero lost ETW buffers or events. These results qualify trace integrity, not performance: the study reports no frame intervals, P99 estimate, or intervention effect. The contribution is an auditable, human-in-the-loop pattern for trustworthy agent assistance on a real workstation, including explicit stop conditions when evidence is insufficient.

Automated plant traits recognition from herbarium images is essential for plant sciences, yet remains challenging because background elements (e.g., textual labels, mounting artifacts, and color charts) can introduce shortcut learning, leading models to rely on spurious non-plant cues rather than plant morphology. This bias degrades both generalization and interpretability. In this paper, we introduce AT-ViT, a dual-branch Vision Transformer that jointly encodes raw herbarium scans and their segmented-derived counterparts via a multi-scale, multi-view cross-attention fusion scheme. AT-ViT further incorporates a mask-guided patch weighting mechanism that amplifies plant-relevant regions and attenuates background-driven features. By learning from the original scans while being guided by segmentation masks through the mask-guided patch reweighting mechanism, the model is encouraged to focus on plant organs and learn plant-centric representations more effectively. Across multiple trait classification tasks (e.g., leaf base shape, thorns), AT-ViT delivers consistent accuracy gains, improves attention localization on plant regions, and exhibits increased robustness under synthetic background perturbations. Specifically, AT-ViT substantially improves spatial attention grounding, boosting plant-region alignment (Avg IoU_p: +15.66 to +18.03 pp) while reducing background overlap (Avg IoU_b: -27.92 to -31.02 pp) relative to CrossViT, and remains markedly more robust to background perturbations, outperforming ResNet101 by up to +32.32 accuracy points and CrossViT by up to +5.07 points under background-noise conditions.

Deploying autonomous systems in safety-critical domains demands guaranteed robustness against physically plausible geometric perturbations rather than abstract pixel-wise noise. In vision-based navigation and autonomous landing, machine learning components require rigorous validation under dynamic operational conditions such as camera rotations and lighting shifts. Extending findings on the failure of first-order spatial attacks in classification, we show that standard gradient-based heuristics (e.g. APGD) similarly fail on for pose estimation, often performing worse than a simple random sampling baseline. To overcome these optimization bottlenecks, we reformulate pose estimation robustness within the framework of Global Lipschitzian Optimization (GLO). We argue that GLO offers a principled approach to robust validation, effectively localizing global optima with strong theoretical convergence guarantees. We evaluate this framework on a YOLOv8-Pose keypoint detector with a Perspective-n-Point (PnP) solver against rotation and contrast. In our evaluations, GLO successfully isolates critical failure modes where position deviations exceed safe operational limits, while rapidly pruning the search space by over 80%. To the best of our knowledge, this is the first study to extend geometric robustness validation to continuous keypoint regression and deep object detection, establishing a practical step toward certifying robust autonomous perception.

Integrated sensing and communication (ISAC) has been identified as one primary usage scenario in the sixth-generation (6G) network. While techniques to preserve information privacy, such as cryptography, have been widely investigated, how to preserve sensing privacy is still an open problem in the literature. This paper makes an early attempt to tackle the above issue. Specifically, we consider a localization system consisting of a multi-antenna transmitter, termed Alice, a single-antenna legitimate receiver, termed Bob, and a single-antenna illegitimate receiver, termed Eve. To allow Bob to estimate Alice's angle-of-departure (AOD) but prevent Eve from performing this task based on Alice's signals, this paper proposes a novel antenna selection and permutation based transmission strategy for Alice. Under this scheme, Alice carefully selects a subset of antennas and permutes their indices to establish a specific pilot-antenna mapping for transmission. Similar to cryptography for information privacy, such a mapping will serve as the secret key to preserve localization privacy. In the special case without noise at Bob and Eve, we manage to find out all the antenna selection and permutation solutions such that with this key (knowledge about the exact pilot-antenna mapping), Bob can uniquely estimate Alice's AOD, while without this key, Eve can estimate multiple AODs of Alice that can lead to its received signals. In the noisy case, numerical results are provided to show that our scheme can confuse Eve to make inaccurate AOD estimation as well.

Pathology foundation models (PFMs) are increasingly used as general-purpose backbones, yet existing benchmarks cannot systematically diagnose their whole-slide cellular representation capabilities, including the decodability of cell-type information and the transferability of such information across tissue sections, datasets, and anatomical organs. We introduce CellPath-Bench, a cellular-resolution benchmark that evaluates frozen PFMs themselves. Following quality control of 52 candidate Xenium datasets, we construct a panel of 25 spatially aligned H&E–Xenium tissue sections spanning 11 organs and 7,079,283 cells, harmonized into fine- and coarse-grained taxonomies. CellPath-Bench samples frozen WSI feature maps at registered nuclear coordinates and evaluates them using standardized multiclass linear probes. Cell Representation Advantage (CRA) measures the within-section advantage of nucleus-anchored representations over patch-level mean pooling, while Cell Representation Transferability (CRT) characterizes the generalization of cell-type decodability across tissue sections, datasets, and organs. We benchmark 30 pathology-specific and general-purpose foundation models through 304,920 runs across spatial readouts, magnifications, taxonomic granularities, and evaluation protocols. The results reveal substantial model-dependent differences in cell-type decodability and its cross-domain generalization, yielding distinct multidimensional capability profiles. CellPath-Bench provides a standardized framework for auditing cellular information in frozen PFM representations.

Shape-constrained and physics-informed learning reports an accuracy cost of enforcing a prior and treats it as a property of the prior. We show it is mostly a property of the free features and the validation split. Let P be the excess risk of restricting a hypothesis class to functions with a shape constraint on features S, and D the excess risk of the ablated model that ignores S. Because a function constant in x_j is both non-decreasing and non-increasing in x_j, the ablated class is contained in the constrained class, so 0 <= P <= D for every risk functional, with no convexity, smoothness, or realizability assumption. Empirically the bound is a sign test: a constrained model must never be beaten by its own ablation. We instantiate it on an ordinal wildfire-severity task (N = 26,681, K = 3) with hard monotone constraints on four meteorological drivers, coordinates left free, and a validation ladder from i.i.d. resampling to 2-degree spatial blocking. Coordinates act as a shield: alone they recover 92.9% of the full model's macro-F1 under spatial blocking, collapsing D from 0.1288 to 0.0427; the same prior costs 0.0473 shielded and 0.3470 unshielded, a ratio of 7.3 with identical physics. Because D is protocol-dependent it does not transfer: coarsening blocks from 1 to 10 degrees drives D from 0.0942 to 0.0050, leaving two configurations unidentifiable a priori. Inversions of the certified nesting bound the pipeline's additive resolution: over 318 comparisons they give a self-calibrating floor of 0.0220 macro-F1, below which no reported price is interpretable, including four cells in our own headline grid. Cost and compliance are independent: the unconstrained model violates the prior at rate 0.48-0.49 while enforcing it costs 0.0473. We give a two-fit screen that rejects unidentifiable experiments before a constrained model is trained.

Agentic large language model (LLM) systems are reshaping scientific workflows in chemistry and drug discovery, but evaluating their open-ended, tool-augmented outputs remains a fundamental bottleneck. The LLM-as-a-Judge paradigm has emerged as a scalable alternative, but existing drug discovery benchmarks deploy LLM judges without validating their alignment with human experts. In this work, we present an LLM-as-a-Judge evaluation framework for ChatInvent, an agentic drug discovery assistant deployed at AstraZeneca, with five contributions. First, we define four output-quality evaluation dimensions—Completeness, Relevancy, Structural Clarity, and Scope Adherence—alongside deterministic Tool Call Correctness checks. Second, we validate the judge through a human alignment study with five expert annotators, comparing Gemini 3.1 Pro, Claude Opus 4.7, GPT-5, and Llama 3.1 70B as candidate judges. Third, we optimize the best-performing judge using few-shot demonstrations of human-annotated examples, improving alignment with the human majority vote from 0.80 to 0.86. Fourth, applying the optimized judge to 70 held-out questions, we surface concrete limitations and find no strong evidence that informal phrasing degrades output quality; it may, however, still be helpful to have the LLM rewrite the original question before querying the agent. Finally, we extend the framework to 38 adversarial questions that are ambiguous, invalid, out-of-scope or ethically sensitive, and show that the agent's refusal behavior is guided by the stated intent of a request. Our framework provides a reusable template for human-aligned evaluation of agentic systems in scientific domains.

Flying in a prescribed formation in an agile manner remains a challenging problem in the field of UAVs, particularly when following highly-demanding trajectories that require flight at platform limits. We address this problem by proposing a novel decentralized approach to formation flight along a given path that integrates formation maintenance into the MPCC framework, allowing UAVs to adapt their progression along complex paths while respecting individual dynamic constraints and maintaining the desired formation. To this end, we introduce a novel reparametrization and synchronization method for dynamic formation geometries together with a decentralized approach to determine the desired positions for the individual UAVs. The proposed approach allows the formation to coordinate high-speed path following without compromising formation integrity. The proposed approach is validated through extensive simulation and real-world experiments involving scenarios with varying complexity of paths and changes of required formation shape on the fly. In comparison to time-parameterized trajectory tracking, we demonstrate improved formation maintenance by 65% in high-speed flight with velocities up to 21 m/s, while achieving comparable times required to reach the goal.

Collaborative perception extends the sensing range of a single vehicle by fusing observations from nearby agents, which improves the robustness of autonomous driving. In realistic deployments, however, the received collaborator messages are often affected by both communication delay and relative-pose noise, which jointly cause stale observations, spatial misalignment, and unstable feature fusion. Existing methods usually address these issues from either the spatial or temporal side, but handling them jointly in a unified and efficient manner remains challenging. In this paper, we propose CoAnchor, an anchor-centric spatio-temporal alignment framework for asynchronous collaborative perception. Instead of directly reasoning on dense BEV features, CoAnchor builds sparse object-level spatio-temporal anchors as a shared interface for pose correction and tightly connects spatial refinement, temporal propagation, and current-time verification within one unified loop, while keeping the overall correction process lightweight. Extensive experiments on both simulated and real-world datasets illustrate that CoAnchor remains competitive under clean settings and improves the robustness under joint delay and pose perturbations with a favorable practical accuracy-efficiency trade-off.

Integrated sensing and communications (ISAC) is a key enabler for uncrewed aerial vehicles (UAVs) in the low-altitude economy. This paper proposes an ISAC waveform that embeds a unique word (UW) into orthogonal chirp division multiplexing (OCDM), termed UW-OCDM, together with corresponding communication reception and cooperative sensing schemes for high-mobility UAV scenarios. For communication, the embedded UW enables timing synchronization and Doppler estimation and compensation without requiring a separate synchronization sequence. A sparse spatio-temporal channel estimation method exploits the common channel support across multiple receive antennas and consecutive UW observations to support reliable data demodulation. For sensing, the deterministic UW serves as a shared prior that allows distributed base stations to construct sensing dictionaries locally without exchanging random payload symbols in real time. A hierarchical multi-target detection and tracking algorithm integrates direct-path interference suppression, kinematic prediction, multi-candidate screening, off-grid refinement, residual verification, and successive interference cancellation for robust localization with reduced search complexity. Simulation results demonstrate reliable communication and localization in highly dynamic UAV scenarios, while the proposed framework retains low-complexity frequency-domain equalization and reduces transmit-reference sharing overhead and multi-static localization complexity.

With the progression in open and disaggregated 6G radio access networks, it is expected that the system will be able to host multi-vendors. In order to host multi-vendors, it is essential that AI-assisted control loops remain safe, verifiable, and auditable under concurrent operator intents and untrusted model inputs. The existing studies address the agentic coordination, formal intent constraints, zero-trust prompt verification and cryptographic accountability in isolation, which leaves pre-realization safety, continuous semantic verification and cross-domain audit incomplete when used individually. In this regard, we propose zero-knowledge auditable control and zero-trust verifiable agentic intent architecture ($Z^2$-ACT), which integrates the aforementioned four primitives across the non-real-time and near-real-time RICs. We encode the typed Intent Contracts as operator goals while the large language model inputs are only admitted after a practical adversarial intent check. The skill sequences in the proposed study are released only when a self-management gate is satisfied while every successful commit is recorded as a binding commitment with a zero-knowledge proof. Our experimental evaluation on public ColO-RAN measurements compares the full architecture against targeted ablations and a conventional reinforcement-learning baseline. A live large language model is used in the non-real-time path to translate operator intents into Intent Contracts; we report translation accuracy, the rate of invalid or hallucinated contracts, non-real-time latency, and behavior under adversarial or misleading intents. Near-real-time control remains trace-driven on the public KPM sequences. Results indicate improved actuation filtering and attack resilience at modest latency and signaling cost inside the near-real-time envelope.

Class-incremental learning is commonly instantiated as a single-model paradigm, where a unified model sequentially adapts to an unbounded stream of sessions. While effective under mild distributional shifts, this formulation becomes strained when successive sessions induce incompatible optimization directions, leading to destructive interference and catastrophic forgetting. We argue that such forgetting reflects a structural limitation of enforcing heterogeneous learning dynamics within a single parameter space. Motivated by social solidarity theory, we propose Socialized Division and Collaboration (SDC) as a reformulation of continual learning that decomposes session learning across specialized models in response to optimization conflicts, while enabling coordinated collaboration. To support this formulation with a principled allocation mechanism, we introduce an energy-based session-model compatibility criterion grounded in Helmholtz free energy, which guides adaptive session allocation and model evolution under conflicting objectives. This framework integrates session assignment, model evolution, and collaborative inference into a unified pipeline, offering an alternative to monolithic continual learning formulations and highlighting a broader design principle for learning under persistent optimization conflicts.

Conventional in-distribution evaluation can overestimate robustness when training and test data share recurring task-specific patterns or surface cues. This risk is especially relevant in social-engineering fraud detection, where attackers can preserve malicious intent while changing the scenario, impersonated entity, or wording. We study this problem as scenario-level out-of-distribution (SL-OOD) detection for SMS and voice phishing, where entire attack scenarios are held out from training while the label space remains fixed. This setting tests whether models can generalize to unseen attack scenarios using decision-relevant evidence rather than familiar scenario-specific cues. Using this SL-OOD evaluation, we find that high in-distribution performance does not reliably predict held-out robustness across feature-, encoder-, and decoder-based baselines. We interpret this gap as scenario memorization: reliance on recurring scenario-specific lexical or entity cues rather than decision-relevant evidence. We propose ECoG, an evidence-consistent generative framework that combines evidence-span supervision with a rationale-label consistency objective during training. On the 0.5B decoder, relative to the same backbone trained without consistency regularization, ECoG raises Macro-F1 on OOD challenging instances by 3.22 points, reduces the share of predictions whose generated rationale supports the opposite label by 4.22 points, and increases token-level overlap with reference evidence spans by 8.38 points; the reduction in prediction-rationale inconsistency is consistent across four decoder backbones. These results suggest that compact generative detectors can benefit from evidence supervision and rationale-label consistency under social-engineering shift.

Geospatial representation learning from satellite imagery is a fundamental problem for large-scale urban analysis and real-world applications. Despite recent advances, current methods struggle with cross-region generalization and semantic interpretability due to their reliance on region-specific auxiliary data and the neglect of semantic alignment within multi-temporal urban imagery. Therefore, we present CoST, a novel \underline{Co}ntrastive-based \underline{S}patial-\underline{T}emporal framework that aligns spatial context with multi-temporal semantics to extract universal geographic regularities shared across regions. Specifically, CoST explicitly models spatial correlations to capture transferable geographic structures and exploits multi-year urban change semantics to align learned representations with high-level geo-semantics. Extensive experiments demonstrate that CoST consistently achieves superior performance across various downstream tasks and in unseen scenario, yielding an average relative gain of 8.7% over the strongest competing methods across eight city-indicator settings. The code is available in \href{https://github.com/Arandinglv/CoST}{this repo}.

Real-world social relationships are not uniformly supportive. Information through hostile connections can increase resistance, anxiety, or misinformation rather than adoption. Classical models such as Independent Cascade and Linear Threshold, together with Influence Maximization (IM), which maximizes spread from a limited seed set, treat activation as discrete and irreversible. Its counterpart, Influence Minimization (Inf-Min), limits undesirable spread but similarly relies on simplified activation assumptions. Signed extensions incorporate polarity but largely retain this irreversibility, leaving no room for beliefs to weaken, reverse, or recover under competing influence. Moreover, both objectives typically treat individuals uniformly, without accounting for differences in vulnerability or prioritizing protection of those most at risk. We introduce RASH, a signed, susceptibility-aware diffusion model in which node awareness is continuous, bounded, and non-monotonic, and prove that despite this added expressiveness it remains monotone and {\gamma}-weakly submodular where only positive or negative edges exist, preserving tractable greedy approximation guarantees where strict submodularity provably fails. Building on RASH, we formulate Harm Minimization (HM), which maximizes aggregate reach while minimizing the awareness shortfall (harm). We prove HM is NP-hard, yet its harm-reduction formulation inherits the same monotonicity and weak-submodularity structure, admitting a greedy algorithm with a bounded approximation ratio. Across six structurally diverse signed networks, RASH is the only diffusion model tested to our knowledge that ever allows awareness to reverse after activation, letting sustained discouraging influence drive awareness from positive toward negative, and HM achieves the highest harm reduction of any method evaluated, including its own boundary cases (IM and Inf-Min)

The transport of dangerous goods by sea is a high-consequence activity governed by the International Maritime Dangerous Goods (IMDG) Code, a complex regulatory framework where errors in classification, packaging, stowage, or segregation can result in fire, explosion, toxic release, or loss of life or vessel. Correct compliance requires accurately interpreting hundreds of pages of interacting provisions, updated on a two-year amendment cycle. Practitioners increasingly use Large Language Models (LLMs) as decision-support tools, yet no systematic evaluation exists of whether they can reliably interpret IMDG requirements for safety-critical use. This paper introduces DGEval, the first benchmark for evaluating LLM knowledge of IMDG Amendment 42-24. Built from expert-written questions on a commercial e-learning platform and structured lookups from the Dangerous Goods List (DGL), it comprises 1,678 questions across multiple-choice, open-ended, DGL lookup, and regulatory identification tasks. We evaluate 13 models from six providers across multiple thinking configurations, including one maritime domain-specific fine-tuned model, and test the effect of web search. Although the best-performing model exceeds the human practitioner baseline on multiple-choice questions, all models are weakest in the operationally safety-critical areas of stowage, segregation, and regulatory recall. These results indicate that LLMs may support compliance tasks, particularly structured DGL lookups with web search, but unreliability in operational areas and regulatory-text recall means human oversight and authoritative source verification remain necessary before deployment in any safety-critical context. DGEval is designed as a safety assurance instrument to be applied continuously as models evolve, not as a settled characterisation of current capability.

Long-horizon manipulation tasks are often only partially ordered. For example, when assembling an electronic device, the battery and circuit board may be installed in either order, but both must be in place before the enclosure is closed. Recovering such dependencies enables robots to flexibly reorder subtasks while preserving task validity. Existing approaches typically infer task structure from human demonstrations using both temporal and symbolic supervision. However, symbolic predicates require explicit grounding, which is difficult to obtain in realistic settings. In this work, we present an approach for extracting task dependency structures from demonstrations using only simple kinematic graphs and distributions over relative object poses. From these representations, our method estimates pairwise task-step-dependency probabilities and uses them to initialize the edge weights of a precedence graph. We then introduce a filtering pipeline that converts this graph of probability estimates into the final task dependency graph. We evaluate our approach on an existing benchmark and on a new dataset comprising longer tasks with more complex dependencies. We find that our method recovers more accurate task structures from fewer demonstrations than the baselines. Finally, we demonstrate that the inferred graphs can be used to generate multiple valid robotic execution orders for the same task.

Satellites are becoming programmable computing platforms capable of running increasingly demanding AI workloads. This shift raises a systems problem: how can AI services remain deployable, manageable, and recoverable after launch when compute capacity, connectivity, energy, and thermal headroom vary over orbital time? This paper develops a systems vision for AI infrastructure in space. We define it as the systems layer that manages AI capabilities across spacecraft, orbital networks, ground stations, and cloud backends, while treating orbital and physical state as part of the resource model. We synthesize relevant foundations from terrestrial AI infrastructure, satellite networking, and satellite edge computing, and examine the physical constraints that directly shape system design. We further ground this vision in three in-orbit case studies spanning the node, platform, and service levels. Telemetry from BUPT-1 satellite shows that usable compute capacity is bounded by thermal and energy envelopes. SateLight on BUPT-2 satellite reduces application-update transmission latency by 56.54% on average and up to 91.18%, with 100% update correctness. A stateful VLM serving case further shows that thermal interruptions make execution-state recovery a first-class systems problem. These observations motivate a research agenda for space-native resource management, lifecycle support, and sustained AI service across space and ground.

Vehicle-to-Everything (V2X) cooperation enables beyond-line-of-sight perception, mitigating occlusions in single-vehicle sensing. However, existing V2X benchmarks provide limited support for closed-loop evaluation and language-grounded supervision, hindering the development of vision-language models (VLMs) for end-to-end cooperative driving. To address these limitations, we introduce V2XBench, a simulation platform featuring synchronized ego–roadside sensing and closed-loop evaluation, together with Chat-V2XBench, a progressively structured VQA dataset for cooperative reasoning. Building upon this benchmark infrastructure, we propose AURORA, an end-to-end cooperative driving framework. Equipped with a dual-view perception architecture, AURORA mitigates spatial and semantic discrepancies across ego and roadside viewpoints through a query-level Cross-View Query Alignment and Fusion (CQAF) module. Leveraging the resulting unified tokens, a LoRA-adapted VLM bridges semantic reasoning and generative trajectory planning. Extensive closed-loop evaluations on V2XBench demonstrate that AURORA achieves state-of-the-art performance in heavily occluded scenarios, with a Route Completion rate of 98.21% and a Driving Score of 76.02, while requiring low roadside communication bandwidth. Ultimately, this work pioneers an extensible V2X–VLM paradigm, paving the way for next-generation cooperative autonomous driving.

We present PhysCaP, a Physics-Informed Code-as-Policy agent system for active perception in robotic manipulation. While vision-language-action policies excel at imitating demonstrations, they rely on passive observation and fail to infer latent physical properties critical for manipulation. PhysCaP augments code-as-policy frameworks with a physics-informed exploration layer that enables explicit information-seeking through interaction. Our method introduces training-free physical property extraction modules that estimate object mass and stiffness from robot proprioception without additional sensors. To balance exploration costs and the efficiency of information obtained, PhysCaP employs a multi-agent design: a Planner that decides when to explore and when to stop, and a Prioritizer that filters implausible interactions and ranks the remainder using a heuristic priority score, enabling efficient, targeted exploration. We evaluate PhysCaP on three real-world tabletop manipulation tasks and a simulated task in LIBERO. The results show that existing passive and naive interactive baselines either fail when physical properties are hidden or over-explore, whereas PhysCaP achieves comparable performance with fewer interactions and reduced execution time. Ablation studies further validate the effectiveness of the proposed physical property extraction modules. Project page: https://physcap.github.io

Video multimodal large language models have advanced significantly, yet fine-grained motion-temporal understanding remains fragile. The core bottleneck is not only sparse frame sampling, but also the lack of a complete temporal modeling pipeline for explicitly representing frame-to-frame change, enabling appearance-motion interaction, and optimizing temporal direction sensitivity. We propose COMET, a temporally grounded framework that systematically strengthens video MLLMs through explicit temporal representation, appearance-motion fusion, and direction-aware optimization. Architecturally, COMET introduces a temporal motion branch built on Taylor frame differences and injects its motion evidence into the appearance stream via temporal attention bias-enhanced cross-attention. For optimization, COMET combines temporal prior distillation with a forward-reverse TC-GRPO stage that turns temporal order into a direct learning signal and strengthens the model's use of directional motion patterns encoded by the temporal motion branch. The method achieves consistent overall improvements with a pronounced motion-temporal bias: on Qwen3-VL-8B, action-centric tasks (STAR, SSv2) improve by 4.9% on average, temporal reasoning tasks (NExT-QA, CLEVRER, LLaVA-178K) by 2.1% over BL-GRPO, while static perception tasks (PerceptionTest) remain on par. The same gain pattern also transfers to InternVL2.5-8B, indicating that COMET generalizes across model families.

LLM agents are emerging as an important paradigm for real-world tasks that require reasoning, tool use, and sequential decision-making. As these agents operate over longer horizons, runtime intervention offers a way to improve reliability without retraining the underlying actor. Effective intervention must provide a useful direction for recovery besides a warning. Existing approaches often rely on an expert solver or a critic that generates task-specific corrections, incurring either the cost of another capable solver or the capacity demands of a task-capable critic. We introduce Comparison-Only Tiny Advisor (COTA) for constructive runtime intervention, which reduces the learned intervention role to local action comparison. A lightweight comparator judges the actor's proposal against available alternatives, and preferred alternatives are returned as non-binding advice for replanning. The comparator is trained from same-prefix counterfactual branches. Across WebShop, ALFWorld, and tau^3-Retail with three LLM actors, COTA instantiated with a 0.5B comparator consistently improves the original actor and achieves the strongest overall performance–cost trade-off among the compared methods. These results suggest that effective runtime intervention need not itself be a task-solving problem: the intervention role can be separated from task solving and handled by a lightweight model specialized for local comparison.