Papers

Image-based dietary assessment promises to replace costly, bias-prone manual recalls, but portion estimation remains a major blocker. Multimodal LLMs (MLLMs) recognize a wide range of foods zero-shot in uncontrolled photos, yet they are weak at portion estimation – a gap we measure across the current frontier (Gemini, GPT, and Claude flagships alike). We present a method that enhances a frozen, commercial MLLM with an accurate portion head: a small geometry-enhanced network on a frozen DINOv2 backbone with a structured softmax-ownership volume, consuming the MLLM's per-food name, bounding box, and density range – no depth sensor, no MLLM fine-tuning. Evaluated fully open-vocabulary on three real-world benchmarks, the head cuts per-food portion error by 33-41% relative to the MLLM alone, outperforms every flagship MLLM's direct estimates, and surpasses each benchmark's originally published image-only model at its own reported metric.

AI-driven algorithms and automated tools are increasingly embedded in the correctional landscape, shaping parole eligibility,release decisions, and surveillance. These tools are also often framed as objective, inevitable solutions to inefficiency andbias. Yet, these computational systems are rarely designed with input from justice-impacted individuals, which means theymight fail to address the real needs of incarcerated people. To address this gap, we surveyed 31 formerly incarcerated peopleabout their parole experiences and their visions for technologies that could support parole preparation. Contrary to dominantassumptions, participants did not imagine computational tools as instruments to dismantle the prison system, but as resourcesfor navigating power: translating complex parole concepts into culturally familiar terms, documenting personal transformationin board-legible ways, and recognizing the often-invisible labor of families. We argue that these imaginaries point towardtechnologies designed for strategic agency, where tools help incarcerated individuals and their families build the capacity tonavigate existing power structures in pursuit of freedom. We conclude by reframing computational tools for parole awayfrom surveillance and toward human-centered systems that support people in navigating carceral power structures.

Oversampling is widely used to address class imbalance in tabular classification, but existing methods can distort the feature importance ranking underlying model explanations. Although recent studies have quantified this distortion by comparing real and synthetic data, none have actively sought to prevent it. In this paper, we introduce Kendall-constrained Importance-Preserving Oversampling (K-IPO), a generator-agnostic, "generate-then-select" framework that preserves the original data's feature importance ranking during augmentation. K-IPO iteratively generates minority-class candidates and accepts them only if their inclusion maintains a user-defined minimum Kendall's tau ({\tau}) correlation with the reference ranking. Optionally, stricter constraints can be applied to the highest-ranked features. We evaluated K-IPO on 20 imbalanced binary classification datasets using three classifiers and multiple explanation methods. In most cases, K-IPO achieved the best or tied-best results in feature importance preservation, explanation consistency, and class separability. It also generally improved predictive performance while maintaining competitive computational overhead.

Software configuration tuning is crucial for optimising system performance, and various optimisers have emerged over the last decade. Yet, the time required during the tuning process may vary across systems. In some systems (e.g., PostgreSQL), it may take a few minutes to measure a configuration, whereas in some others (e.g., MariaDB), it can take several hours. Moreover, even within the same system, users may have varying budgets and preferred settings. This naturally raises a question – Given a budget level, which optimiser is the best choice for SE practitioners? This matters because optimisers usually have their own “comfort zone” and may perform very differently under distinct budgets. In this paper, we aim to answer this question. We systematically evaluate eight well-established optimisers across 22 configurable systems under varying budget levels. We find that, unsurprisingly, model-based optimisers (e.g., SMAC) are well-suited under tight budgets, and model-free optimisers (e.g., GAs) become superior with more generous budgets. However, interestingly, there is one optimiser, FLASH, that performs consistently well on most systems regardless of budgets. We lastly investigate the reasons behind this phenomenon and find that many systems possess good local optima (with large basins of attraction), allowing greedy optimisers (e.g., FLASH) to achieve strong performance. Source code, data, and supplementary materials of this work are available at https://anonymous.4open.science/r/Config-W2W-98B2.

Edmund C. Berkeley is usually remembered as a mediator between symbolic logic and early computing, yet that standard description understates the scope of his work. This paper argues for a stronger reading: Berkeley should also be understood as an early theorist of embodied machine intelligence. Across Berkeley's major writings on symbolic logic, machine intelligence, living robots, and Squee, intelligence appears not as disembodied symbol manipulation alone but as the organized coordination of sensing, storage, calculation, control, state, and action in physically realized machines. The paper's first contribution is interpretive: it reconstructs Berkeley as a thinker of machine architecture, temporally extended behavior, and environment-coupled control. Its second contribution is comparative: it reads Berkeley alongside David L. Heiserman to recover a shared descriptive scheme centered on sensing, state or memory, control, action, and adaptation. Its third contribution is critical: it uses that scheme to assess current embodied-AI discourse. The broader claim is that contemporary LLM-centered robotics often demonstrates impressive capability without an equally explicit account of persistence, recoverability, maintenance, and structured modification of conduct through experience.

Chat models sometimes commit to an answer and then produce reasoning that justifies it rather than deriving it – even when the answer contradicts a task premise. We study a minimal probe: "I want to wash my car. The car wash is 100 meters away. Should I walk or drive?" Only drive works (the car must be at the car wash), yet models overwhelmingly recommend walking. (1) Behavioral reproduction: on Qwen3-8B across five system-prompt conditions (210 rollouts), the wrong commitment occurs in 85-100% of sampled rollouts per condition and 100% of greedy rollouts, in both thinking and non-thinking modes; a 4,096-token thinking budget does not repair it. (2) Preliminary activation-level evidence: probing hidden states with a pretrained, training-free activation oracle (no task-specific probe training) at positions before the answer text is emitted, "walk" read-outs exceed a neutral-context baseline (68% vs. 17%; walk-committing rollouts p=.005, drive-committing rollouts p=.005, Fisher exact) – notably, rollouts that eventually answer drive also read as walk-leaning before commitment (5/6). The oracle's default on unrelated content is "drive" (83%), so the read-outs are not lexical bias; stratifying by literal walk/drive occurrence shows they are not text recovery either (spans containing "drive" still read out walk; in balanced lexical fields, per-rollout walk-majorities beat a per-prompt neutral baseline 15/22 vs. 1/8, p=.01; drive-committing rollouts 6/6, p=.002). Samples are small and the within-rollout positional gradient is not significant (p=.34); we frame these results as preliminary. (3) Methodological: with fixed oracle, activations, and positions, question wording alone moves a positive control from 2/16 (open question) to 11/16 (closed); negative oracle results are uninterpretable without per-wording positive controls.

Interpretability methods for neural network activations span a wide cost spectrum, from cheap, training-free techniques (such as linear probes, PCA, SVD) to more expensive training-based ones (such as SAEs and activation oracles). Training-based methods are typically more powerful, in part because they leverage large activation datasets during training. This raises a natural question - do they actually surface insights that go beyond what is recoverable from the training dataset itself? To address this, we equip an LLM agent with a vector database of activations paired with their textual contexts, along with tools for manipulating activations - projecting out directions in latent space, computing activation differences and averages. The agent iteratively queries the database, forms hypotheses from the retrieved samples, and validates them by constructing linear probes. We call this method HARP, for Hypothesis-driven Agentic Retrieval and Probing. Despite not involving any training, HARP outperforms both activation oracles and SAE-based agents on concept discovery, concept detection, model steering, and secret elicitation. The training-free design also makes HARP substantially cheaper and more flexible: new datasets can be indexed on demand whenever existing ones prove insufficient. More broadly, our results suggest that current training-based methods do not yet extract insights beyond their training data, and motivate benchmarks that explicitly require interpretability methods to demonstrate such insights. We release our code at https://github.com/SriramB-98/HARP

For a second time, the android robot Andrea was set up at a public museum in Germany for six consecutive days to have conversations with visitors, fully autonomously. Building on previously gathered qualitative results, the robot was now capable of engaging in multi-lingual conversation with the visitors about the museum context. The robot was prepared with context information about the museum in general and its surrounding exhibits this time. The robot featured a slightly artificial sounding voice that was previously evaluated as congruent with its gender-ambiguous but very humanlike design. Three experimental conditions were implemented, in which either (1) the robot simulated no emotions, (2) the robots emotions were determined by ChatGPT 4.1, or (3) the WASABI emotion simulation architecture simulated the robot's emotion dynamics. An extended version of the TAM2 questionnaire was employed to let 73 visitors report on several factors of their opinion about the android robot Andrea after having experienced it. In result, the statistical analysis suggests that these first two approaches to implementing emotions into the chat architecture of our android robot Andrea did not yield any positive effects on the subjective evaluations by the visitors and were not detectable on a conscious level.

It has long been recognized that humans have the ability to switch between fast, reactive decision-making and slower, deliberative planning. In this paper, we study the question of how to learn this ability, known as meta-reasoning, in artificial agents. We model reactive decision-making as a policy that directly maps state observations to actions. Such policies can be trained with reinforcement learning (RL) or imitation learning, but may generalize poorly outside of their training distribution. Alternatively, model-based decision-time planning is more likely to produce good actions across a broader set of states but requires additional computation time, which delays acting. In this work, we introduce an RL method for training a meta-reasoning policy that allocates computation by conditioning on a reactive-policy uncertainty score. This score enables it to predict when the reactive policy is likely to perform poorly and when planning is needed. We conduct an empirical study on motion planning and navigation environments, showing that this design enables the meta-reasoning policy to learn when the reactive policy provides a good-enough action versus when decision-time planning is needed. Additionally, we show that our design enables the meta-agent to shift toward fully reactive control as the reactive policy improves.

Artificial intelligence (AI) systems are now ubiquitous across domains such as security, finance, healthcare, consumer technology, and large-scale cloud services, where they process massive volumes of data and make consequential decisions daily. This widespread adoption has created a broad attack surface through which adversaries can manipulate, evade, extract information from, or otherwise subvert deployed models. Depending on system design and exposure, attackers may have very different forms of access: some observe only final decisions, while others receive confidence scores, intermediate representations, or even full model parameters. While previous surveys typically organize evasion attacks into white-box, gray-box, and black-box categories based on the attacker's knowledge of model internals (architecture, parameters, gradients), this taxonomy often conflates different deployment scenarios that provide vastly different output signals, all labeled as “black-box” despite enabling fundamentally different attack strategies. Understanding how evasion attack strategies adapt to the specific information signals returned by deployed systems is critical for organizations making procurement and deployment decisions. To address this gap, we introduce the Signal-based Model Access Risk Taxonomy (SMART), a deployment-oriented framework that classifies attacker access according to the nature and richness of the information signals available from deployed AI systems. Using this taxonomy, we provide a structured overview of evasion attacks across progressively richer levels of information exposure, highlighting how deployment interfaces influence attack capabilities and informing more secure AI deployment and procurement decisions.

Current benchmarks for language models primarily evaluate execution on fully specified tasks. However, real user tasks are often ambiguous. Users arrive with incomplete, exploratory, or even inconsistent goals, requiring the assistant to first determine the intended task before carrying it out. We study this problem as task alignment: the ability to align with a user on their intended task. We introduce a general framework for converting specified tasks into underspecified interactions, formalized as a POMDP in which the model must infer a latent task from partial and evolving user intent. We validate our user simulator post hoc with a human user study. Across shopping, coding, and professional work settings, we find that while models often perform well once the task is specified, models still struggle with task alignment: current models act prematurely, interact ineffectively, and fail to resolve ambiguous requests. Models on average recover the user's intended task only 22-32% of the time under ambiguity. In a human study in the same setting, humans reach 48%, outperforming all evaluated models. We show that post-training with supervised fine-tuning and reinforcement learning improves task alignment, but models still lag behind humans in resolving uncertainty through interaction. Together, our results suggest that current models still lack key interaction abilities required for reliable agency.

Unified Multimodal Large Language Models (MLLMs) offer a promising paradigm for unifying visual understanding and generation, yet they still struggle to follow complex spatial instructions and logical constraints in controllable image generation. To address this gap, we present ATLAS, a unified framework that equips MLLMs with a human-like "Think, Plan, and Paint" paradigm. We adopt layout as the shared representation that connects the three stages, enabling the model to reason about spatial requirements, plan explicit object arrangements, and render the final image. We further improve plan-to-image fidelity with reinforcement-learning-based layout alignment. We instantiate ATLAS at 7B and 80B scales, achieving state-of-the-art performance among MLLMs on image generation benchmarks and an average 65.31% improvement over existing layout-based unified MLLMs. On spatially related tasks, ATLAS obtains an average 23.06% gain over the base models. Through the same layout interface, ATLAS also supports instruction-guided editing and multimodal grounding. We further introduce ATLAS-Reasoning, a benchmark for evaluating generation under complex spatial instructions.

Large-scale AI datacenter platforms comprise thousands of heterogeneous hardware components whose validation requires comprehensive fault injection test plans. Today these plans are authored manually: engineers review hardware self-healing validation documents and bills of materials, enumerate failure modes per field-replaceable unit, and produce flat lists of single-layer test cases. This process is labor-intensive, error-prone, and dependent on institutional knowledge; coverage gaps surface late, traceability to source specifications is implicit, and the effort is largely repeated per platform. This paper presents a generative AI multi-agent architecture that automates the generation of structured hardware validation test plans from two canonical inputs: self-healing validation documents, which enumerate known failure modes and their detection and remediation behaviors per field-replaceable unit, and component Bills of Material. An ingestion agent normalizes heterogeneous inputs into a canonical representation; a classification agent maps components to functional domains via contextual reasoning over part descriptions and sub-category hierarchies; and a generation agent synthesizes test cases by combining normalized failure modes with domain-classified data, filling gaps and producing edge cases. The output conforms to a standardized schema for direct import into internal validation software. Evaluated on two production platforms against manual baselines, the framework achieves coverage expansions of 74.2% and 51.4%, cutting authoring from days to hours. It yields fully traceable mappings from each test case to its source specification, and its multi-agent decomposition is portable across platform generations. Automated and expert evaluations confirm 100% extraction fidelity and high acceptance of new scenarios, validating the framework as a robust human-in-the-loop force multiplier.

An agent system's execution traces record how it fails, and procedures that improve such a system without changing model weights (trajectory selection, prompt and workflow optimization, runtime monitoring) read these traces for feedback. Yet raw traces are a poor medium for accumulating that feedback: long, instance-specific, and lacking a stable vocabulary for recurring failures. We argue that an agent system should instead maintain an explicit representation of how it fails, induced from its own behavior and reusable wherever failure feedback is needed. AdaMAST builds this representation by converting a target system's traces into a compact, evidence-grounded failure taxonomy: named failure codes organized along three fixed axes (system-level, role-specific, and domain-specific), with every name, definition, and evidence pattern induced from the traces; no code is hand-authored, no trace human-annotated. The taxonomy is not merely a post-hoc diagnostic but a shared feedback interface, improving agents in three ways. In agent-system search, taxonomy-coded diagnoses of failed candidates outperform free-form reflection on all five benchmarks we test. At runtime, taxonomy feedback raises SWE-agent's resolution on SWE-bench Verified Mini from 60% with free-text reflection to 70%, and improves Claude Code from 64.0% to 70.7% as a runtime skill. In trajectory selection, AdaMAST-Judge, a verifier built on the induced codes, improves best-of-5 accuracy on Terminal-Bench 2.0 by 8-15 points over Pass@1. The vocabulary itself is compact (an order-of-magnitude compression that preserves trace distinctions), human-faithful (matching expert failure annotations more closely than a hand-crafted reference vocabulary), and adaptive (taxonomies induced for different domains share few codes). Adaptive failure taxonomies close the loop between the traces agents produce and the procedures that improve them.

Interactive theorem proving (ITP) underpins program verification and formalized mathematics, but its manual effort limits scalability. LLM-based proof agents promise to ease this effort, but their heavy token consumption and API cost remain a major obstacle. We trace this cost to a shared root: current agents operate on serialized concrete syntax, emitting proofs as source text and recovering proof states through separate, line-number-based queries, so every edit shifts later lines and forces repeated relocation of errors and states. This same dependence on concrete syntax also blocks adoption of Minilang, a recent proof language that reaches SOTA on LLM-based proving but is too new for LLMs' training corpora. We address both problems by lifting the agent off source text and onto the abstract syntax tree (AST): the model supplies proofs as JSON representations of Minilang's AST – native to tool-calling LLMs – and drives the prover through a tree-edit model that fuses proof operations and states into one proof tree, so each operation carries its own subgoal's state, readable directly off the tree. We realize this design in Agent over AST (AoA). Against Amazon's Isabelle Agent on miniF2F and NTP4VC-Pearl common success sets, AoA cuts API cost by 2.3–4.7x (normalized input-cache accounting), uses 2.9–6.9x fewer tokens and 3.9–8.9x fewer tool calls, and finishes 1.4–2.0x faster – while also solving far more problems on the harder verification benchmark.

Robotic navigation in unstructured environments requires robust situational awareness to safely traverse hazards such as steep slopes and rocky terrain. To address this challenge, perception systems increasingly rely on multimodal sensor fusion. Specifically, integrating thermal imagery with standard optical and depth sensors enhances terrain differentiation, directly improving the reliability of mapping algorithms. This paper presents PRISM, a multimodal perception system for terrain mapping in unstructured settings. PRISM leverages a custom sensor suite to capture aligned RGB, depth, and thermal (RGB-D-T) imagery. At its core is OmniUnet, a novel vision transformer-based network specifically designed for multimodal semantic terrain segmentation. We validated the proposed system using two newly annotated datasets (BASEPROD and LAENTIEC) and demonstrate its real-world applicability through physical field experiments. Deployed on a resource-constrained embedded computer, PRISM efficiently generates traversability maps that directly enable autonomous navigation via a rover's Guidance, Navigation, and Control (GNC) subsystem.

This paper presents a unified quantum-classical hybrid framework for multi-horizon time-series forecasting, introducing two model variants Quantum Reservoir Forecaster (QRC-F) and Variational Quantum Forecaster (VQF-F). The proposed framework investigates the complexity-fidelity trade-off of quantum forecasting under near-term NISQ hardware constraints. Continuous time-series signals are transformed into binary representations through uniform quantization and encoded into quantum states using angle encoding with parameterized RY rotation gates. Cross-channel entanglement layers capture dependencies among multiple variables. QRC-F utilizes a fixed random unitary quantum reservoir for stable, gradient-free temporal feature extraction, whereas VQF-F employs a trainable variational quantum circuit optimized through the parameter-shift rule to learn temporal and inter-variable patterns from Pauli expectation values. Both models replace computationally expensive quadratic self-attention with efficient linear transformations, reducing parameter complexity. A shared MIMO-based multi-horizon prediction head simultaneously generates forecasts across multiple horizons, avoiding error accumulation in recursive forecasting. Experimental evaluations on benchmark datasets, including ETTh1, ETTh2, ETTm1, ETTm2, Weather, electricity, and exchange-rate, demonstrate that VQF-F achieves superior training stability and parameter efficiency, while QRC-F provides enhanced robustness and circuit fidelity under quantum noise. The results establish a practical quantum-native forecasting framework with strong potential for deployment on near-term NISQ devices.

Security rating platforms summarize externally observable cyber exposure and are expected to help organizations prioritize remediation. A platform may want to tell an organization how a candidate remediation action would affect its score, but repeatedly exposing exact score responses can reveal information about the hidden scoring engine. We propose a surrogate based approach for remediation score impact prediction that is designed to respect this opacity constraint. The surrogate predicts scores from organization configurations while explicitly representing checkpoint (i.e., a security check) applicability and the observed checkpoint set. A main challenge is that such predictions are not uniformly reliable: they depend on the amount and structure of the observable checkpoint evidence available for a given configuration. To address this, the approach combines applicability-aware surrogate construction, sensitivity analysis under controlled checkpoint restriction, a reliability layer for identifying unstable predictions, and score-impact prediction for supported remediation actions. Explicit modeling of checkpoint applicability is central throughout: it improves score prediction and provides the feature basis used by the reliability layer to identify unstable cases. We evaluate the approach on a real-world dataset of 5,188 organization configurations from a commercial security rating platform. The results show that the applicability-aware surrogate improves score prediction over simpler feature representations. For remediation, the surrogate predicts the score impact of supported actions, while the reliability layer helps identify cases in which these predicted impacts should be interpreted cautiously.

Recent advances in physics-grounded video generation leverage physics simulation as a physical prior to guide video synthesis toward physically plausible outcomes. The simulation process is controlled by physical specifications, which are typically generated by a vision-language model in a single pass. Such one-shot prediction often fails to accurately translate user intent into executable simulations, particularly for fine-grained object dynamics, complex motion trajectories, and temporally structured interactions. In this paper, we propose PhysAgent, a reflective agentic framework that closes the loop among physical program generation, physics simulation, stage-specific verification, and targeted program repair. Beyond improving the control of coupled physical parameters, our framework enables the agent to progressively realize complex trajectories, multi-stage interactions, and precise event outcomes by treating each physical program as an executable hypothesis. In addition, we design a set of physics-control APIs to support more stable and complex motion behaviors. Extensive experiments demonstrate that PhysAgent produces more physically plausible videos, achieves better prompt alignment, and generalizes more effectively across diverse physical scenarios.

Retail demand forecasting remains difficult when demand shifts faster than static forecasting models can be retrained, especially in early demand cycles where newly observed labels are sparse. To address this, this study aims to improve adaptive retail forecasting by proposing a predict-then-correct (PtC) framework that retains a first-stage machine learning (ML) forecast and applies a few-shot continuous contextual bandit correction policy with similar-SKUs augmentation and top-p masked updating. Across Walmart retail data and an exclusive beverage dataset, PtC delivers statistically significant reductions in MAPE, MAE, and RMSE across stable & high volume, stable & low volume, and erratic & intermittent demand patterns, improves average RMSE by 9.52% over the ML-only baseline in the ablation study, and yields lower inventory costs than base-stock, proximal policy optimization, and soft actor-critic policies under the tested lead-time settings. These findings show that online forecast correction can bridge offline demand learning and real-time retail decision-making by adapting to sparse feedback without fully retraining the base forecasting model.

A fundamental intent asymmetry plagues modern 3D asset creation: while state-of-the-art 3D toolchains demand precise, executable parameters, ordinary users typically provide vague, underspecified instructions. Current 3D agents treat this ambiguity as noise, defaulting to blind execution under a single-turn assumption. To address this limitation, we introduce CLARE, a clarification-aware and evolutionary 3D agent that treats intent asymmetry not as an execution error, but as an opportunity for strategic dialogue. By decoupling the generation pipeline into four specialized cognitive roles, CLARE intercepts and resolves underspecified instructions before invoking computationally expensive 3D tools to seamlessly execute tasks across five diverse domains: text-to-3D generation, single-view reconstruction, multi-view reconstruction, point cloud editing, and post-processing. Crucially, rather than relying on rigid manual rules, CLARE self-evolves its clarification policy via simulated multi-turn interactions. By optimizing a Multi-turn Reward, the agent internalizes the delicate balance between interaction efficiency and task completion. To rigorously test this, we construct 3D-Clarify, a comprehensive benchmark comprising 620 interaction scenarios with systematically injected ambiguity, missing information, and mistaken details. CLARE achieves state-of-the-art performance, with 60.40% and 43.34% success rates on single-step and multi-step tasks, respectively, more than doubling existing baselines. Both quantitative and qualitative results demonstrate that proactive clarification is the missing key to robust 3D execution. Code is available at https://github.com/xyzhu1225/CLARE.

Publicly shareable construction-video benchmarks remain scarce, especially for safety-critical hazards that are rare, dangerous to stage, and difficult to release. We study worker under suspended load, a relational hazard that depends on worker-load geometry and temporal persistence rather than object detection alone. We introduce SynthSite, a focused synthetic video benchmark of 55 clips spanning varied load configurations, viewpoints, clutter, occlusions, and surveillance conditions, together with a privacy-aware hybrid generation workflow that supports both publicly shareable benchmark creation and privacy-constrained synthetic video generation. We then ask whether worker appearance can be suppressed without undermining downstream hazard recognition. Under five whole-body privacy conditions, we evaluate worker and load retention, localization stability, and clip-level hazard recognition. We find that structure-preserving obfuscations retain substantially more downstream utility than appearance-smoothing baselines, and that preserving a raw visual reference alone does not guarantee the strongest agreement with human hazard labels. These findings suggest that privacy evaluation for construction safety analytics should assess not only appearance suppression, but also preservation of the geometric cues required for hazard reasoning. Our dataset and code are available at https://huggingface.co/datasets/govtech/SynthSite .

Sensor-based human activity recognition (HAR) has achieved significant progressed in fully supervised learning settings. However, these supervised learning models rely on large amount of labeled data, which require labor-intensive collection and meticulous annotation. To address these challenges, this paper proposes a Joint Embedding Predictive Architecture framework tailored for sensor-based HAR, designed to learn robust and generalizable representations from unlabeled datasets. The proposed framework features an encoder designed to explicitly model both the fine-grained local temporal representations within individual window and the long-term temporal sequence of adjacent windows. Furthermore, we introduce an improved Variance-Invariance-Covariance Regularization (VICReg) objective function that incorporates computationally lightweight norm term to stabilize the JEPA pre-training phase. This term balances variance, invariance and covariance constraints to prevent representation collapse. The proposed HAR-JEPA framework is evaluated using two benchmark continuously performed activity datasets. The results show that high-quality representations are successfully learned by the proposed framework. Furthermore, the representations learned by HAR-JEPA demonstrates superior generalization on minority, high variance transitional activities such as sit-to-stand and sit-to-lie where supervised learning tend to overfit due to limited support.

Machine learning-based intrusion detection systems (IDSs) often suffer from class imbalance and vulnerability to adversarial attacks, leading to degraded detection performance and reduced robustness. This study proposes a TabTransformer framework augmented by the Boundary-Seeking Generative Adversarial Network (BGAN) for flow-based intrusion detection using the CICIDS2017 dataset. BGAN serves a dual purpose by generating synthetic minority-class samples to mitigate data imbalance and producing adversarial samples to evaluate model robustness. Experimental results demonstrate that BGAN augmentation improves TabTransformer's Macro-F1 score from 82.96% to 86.50%, with the largest class-wise improvement observed for Web_Attack (F1 score: 0.29 to 0.61). Robustness evaluation shows that all non-augmented models experienced a 100% Performance Drop Rate (PDR) under adversarial testing, whereas all BGAN-augmented models achieved negative PDR values, indicating improved resilience. Furthermore, the augmented TabTransformer maintained stable and low False Triggered Rate (FTR) values (1.51%-2.92%) across all noise levels, compared with the BGAN-augmented Decision Tree, which reached 49.09% under benign perturbations. These findings demonstrate that BGAN consistently enhances both class balance and adversarial robustness, while the proposed BGAN-TabTransformer framework provides an effective and adaptive intrusion detection solution for adversarial network environments.

Connected and Autonomous Vehicles (CAVs) rely on interconnected software and hardware components, including sensors, Electronic Control Units, in-vehicle infotainment systems, and telematics units, where vulnerabilities can compromise assets, users, and vehicle operations. These vulnerabilities are commonly documented as plain text in the Common Vulnerabilities and Exposures (CVE) database; however, security practitioners require structured information about affected assets, types of weaknesses, and attack behaviors to effectively mitigate the risks from these vulnerabilities. To this end, we evaluate open-weight Large Language Models (LLMs) for generating Structured Threat Information Expression (STIX), a well-known structured format for representing threat information, for CAV-related CVEs. We construct a dataset called CAV-STIXGen that maps CAV vulnerability descriptions to STIX domain objects (SDO), STIX relationship objects (SRO), Common Weakness Enumeration (CWE), and MITRE ATT&CK techniques mappings. Using this dataset, we evaluated 11 open-weight LLMs (4B to 120B parameters) across various prompting strategies and temperatures. Single-model configurations achieve F1 scores of 0.94 for SDO, 0.63 for SRO, and 0.99 for CWE mapping, while complete MITRE ATT&CK mapping remains challenging. In a multi-agent setup, Gemma-4-31B and Codestral-22B achieve F1 scores of 0.91 for SDOs and 0.43 for SROs, respectively. Lastly, we analyze CWE and MITRE ATT&CK co-occurrences to identify recurring threat patterns in the CAV domain, demonstrating how AI-assisted vulnerability-to-STIX translation can automate threat intelligence and prioritize defense in transportation security.

Muon is competitive with AdamW in large-scale pre-training, but its operating regime in reinforcement-learning post-training remains unclear. We map this regime on ALFWorld, a sparse-reward agentic benchmark, using three group-based objectives and Qwen2.5 models from 0.5B to 3B. Under a shared KL and clipping recipe, matched optimizer comparisons and AdamW rate controls trace the usable step-size range. AdamW responds non-monotonically to rate, whereas fan-in Muon remains stable at a more aggressive effective step: at $3 \times 10^{-5}$ it improves late success over an AdamW $10^{-6}$ baseline after correction across rate-metric tests. Its normalized-AUC effect is directionally positive but less uniform; the heuristic-matched lower-rate effect is less consistent, and tuned AdamW nearly matches high-rate Muon at 3B GraphGPO. High-rate Muon applies $3.53 \times$ AdamW's hidden-matrix update RMS; a full-budget RMS-matched control removes the late-success gain. Together, these results identify a recipe-level operating regime in which fan-in Muon supports a more aggressive stable effective step under shared KL and clipping: the margin is largest when optimization headroom remains and contracts near saturation, after AdamW tuning, or under magnitude matching. The scale-matched control ties this spectral effect to Muon's scale convention rather than establishing a universal optimizer ranking. Code is available at https://github.com/x66ccff/verl-muon.

Human vision is a closed loop: gaze is continuously redirected by intermediate hypotheses rather than a single snapshot. Decades of psychophysics and cognitive science have argued that this active observation is essential for a wide range of tasks. Whether today's multimodal large language models (MLLMs) exercise active observation is an empirical question that current vision-language benchmarks do not answer. We introduce ActiveVision, a benchmark that makes active observation measurable for MLLMs, comprising 17 tasks across 3 categories. Tasks are designed to force repeated visual perception rather than a single static description. Frontier MLLMs collapse on ActiveVision: the highest-scoring model we evaluate, GPT-5.5 at the highest exposed reasoning-effort tier, solves only 10.6% of items and scores zero on 11 of the 17 tasks, and even Claude Fable 5, despite topping most reasoning and coding leaderboards, solves just 3.5%, far behind three human participants who average 96.1%. Furthermore, much of the gap persists even when models write and run their own vision code: such code is unreliable on realistic imagery, and catching its failures itself requires the active perception the models lack. Together, these results indicate that current MLLMs lack robust active visual observation, motivating architectures and training objectives that close the perception-reasoning loop.

LLM powered multi-agent systems (MAS) have emerged as a promising paradigm for complex tasks. However, their advantages over single-agent systems (SAS) remain unclear, with performance varying inconsistently across settings. Here, we provide an information bottleneck perspective on elucidating the differences between MAS and SAS. Specifically, our key observation is that a SAS accumulates its full reasoning trace in one shared context, while a MAS uses isolated local contexts connected by bounded relay messages. We show that, under infinite relay bandwidth, any SAS can be simulated by a MAS that transmits the full upstream context. Thus, the nontrivial advantage of MAS arises under bounded relays, where compression introduces a fundamental trade-off: reducing redundant context can improve efficiency, but may also incur loss of task-relevant information. We formalize this trade-off as an information bottleneck controlled by an effective parameter $\beta$, which captures how the balance shifts with model capability, and shows that MAS gains arise when context reduction outweighs relay information loss. We conduct 18 controlled experiments across five benchmarks and three model scales to validate our theoretical studies. We observe that MAS consistently helps when relays are near-sufficient, especially for weaker models. In contrast, MAS gains shrink or reverse when relays incur information loss, especially for stronger models that can already extract useful information from redundant context and thus gain little from compression. Our study shows that multi-agent design is fundamentally an information-bottleneck optimization problem. This perspective explains when bounded inter-agent communication helps or hurts.

Multimodal Scientific Claim Verification (MSCV) requires models to verify scientific claims using visually grounded evidence from papers, including figures, tables, charts, and textual context. However, existing methods often fail because they struggle to locate decisive visual evidence, accurately read structured scientific visuals, and integrate multimodal observations into reliable reasoning. We introduce ToolSciVer, the first tool-augmented framework for MSCV to our knowledge. ToolSciVer equips a VLM with three type-aware visual tools, table row/column focus, chart-to-structure parsing, and high-resolution region zoom, which convert dense scientific visuals into explicit, claim-facing evidence, and trains the policy with Group Relative Policy Optimization (GRPO) under a composite reward of answer correctness, format validity, length control, tool-use efficiency, and tool-validity penalties. Experiments on SciVer and MuSciClaims datasets on five VLMs from three model families (Qwen, InternVL, Gemma) demonstrate that our method achieves superior performance compared to four competitive baselines including prompting-based and RL-based tool-use methods, highlighting the effectiveness of learned, type-aware tool use for scientific claim verification.

AI governance increasingly requires judgments about whether an AI system remains adequately trustworthy over time, whether observed changes are tolerable, and how such judgments should be documented in a transparent and contestable way. Existing approaches remain either too high-level to support lifecycle monitoring and reassessment or too narrowly metric-driven to connect multidimensional trustworthiness evidence with governance decisions. We propose a lightweight methodology centered on trustworthiness level functions: auditable rules that map measured trustworthiness profiles to governance-relevant levels. The methodology separates the underlying trustworthiness evidence from the governance rule used to interpret it and treats that rule as a lifecycle governance object. The rule may remain expert-defined or, when available evidence warrants empirical learning, be approximated by an interpretable candidate model. An AI lifecycle governance procedure embeds this choice in explicit decision gates for determining whether learning should be attempted and whether a learned candidate should become operative. The resulting rule supports lifecycle monitoring through level transitions, boundary margins, and profile drift, with explicit human responsibilities for validation, approval, and reassessment. We illustrate the methodology on synthetic AI lifecycle scenarios involving degradation, shocks, updates, heterogeneous monitoring cadences, and system comparison. Our methodology does not replace expert or legal judgment, but makes the governance interpretation of trustworthiness evidence more explicit, auditable, and contestable over time.