artificial intelligence

The Autonomy Externality: A Welfare-Economic Model of Agentic Artificial Intelligence, Correlated Failure, and Optimal Assurance Policy

Agentic artificial intelligence is shifting automation from prediction and recommendation toward autonomous execution. This transition creates an economic externality that is not captured by conventional firm-level investment models. A firm receives much of the productivity benefit from delegating decisions to an artif...

Methodological approaches to assessing enterprise economic security in the context of digital transformation

The findings emphasize the need to update traditional methodologies to ensure a more accurate and comprehensive assessment of enterprise economic security under conditions of accelerating digital transformation.

Harnessing Artificial Intelligence for the Early Detection and Sustainable Management of the Red Palm Weevil, Rhynchophorus ferrugineus (Olivier)

Functional compatibility as a determinant of persistent neural learning

Functional compatibility is identified, the extent to which incoming learning can coexist with behaviour that must be preserved, as an experimentally manipulable causal determinant of persistence.

PAC-MAN: Perception-Aware CBF-RL for Whole-Body Safety in Humanoid Dodgeball

We present PAC-MAN, a perception-aware CBF-RL framework that couples control-barrier safety with deployment-realistic onboard sensing for whole-body humanoid dodgeball. The deployed policy sees the ball only as segmentation-masked depth from a head-mounted camera, while training-time CBF guidance represents clearance t...

Unlocking Spatial Grounding in Large Audio-Visual Retrieval models

Weak supervision sets a practical regime for audio-visual sound source localization as dense spatial annotations are costly to obtain at scale. The task, however, remains challenging, as models must locate sound sources from temporally aligned audio-visual data without pixel-level supervision. Recent large-scale audio-...

LOCKS: Page-Local Compact Key Summaries for Efficient Long-Context Decoding

LOCKS stores a rank-$r$ spectral summary per page, reconstructs its within-page logits, and selects pages by log-sum-exp mass without reading candidate keys or values, and retains quality furthest under tight budgets on AIME26 and MATH-500.

A Safety-First Gateway Architecture for Trusted Public Health Resource Navigation

Conversational AI can improve access to public health information, but public-facing healthcare applications require safeguards against inappropriate medical guidance and unsupported generation. We present a Safety-First Science Gateway for maternal and child health (MCH) resource navigation that combines large languag...

Does Anthropomorphic Language Impact Public Perceptions of AI?

The results indicate that any immediate effects on opinions of AI are modest, although they leave open the possibility that anthropomorphic language could have an effect in naturalistic settings, or over gradual, continued exposure.

Low-power analogue neural networks with trainable nonlinear connections for continuous control

Inspired by Kolmogorov-Arnold networks, trainable nonlinear functions are placed on the connections, making each physical connection a learnable computational element in physical neural networks.

Is Agent Code Less Maintainable Than Human Code?

It is found that agents are less effective at resolving tasks when building on agent code compared to human code, with task resolve rate drops of up to 13.1%.

Temporal Self-Imitation Learning

Temporal Self-Imitation Learning (TSIL), a reinforcement learning framework that mines temporally efficient successful trajectories generated during learning and converts them into reusable supervision for future policy improvement, is introduced.