Topic
multi-modal
ProMUSE: Progressive Multi-modal Uncertainty-guided Staged Evidential Alzheimer Disease Classification
ProMUSE is a progressive multi-modal uncertainty-guided staged evidential network for Alzheimer disease classification. It uses low-cost clinical data first, then decides when to incorporate expensive MRI or PET imaging based on uncertainty, reducing imaging usage by 50-90% while maintaining accuracy.
OmniMouse Brain Model Trained on 150 Billion Neural Tokens Reveals Unusual Scaling Laws
Researchers trained OmniMouse, a multi-modal, multi-task brain model, on 150 billion neural tokens from 3.1 million mouse visual cortex neurons. The model achieves state-of-the-art performance across neural prediction, behavioral decoding, and neural forecasting. Scaling analysis shows performance improves with more data but gains from increasing model size saturate, contrasting with language and vision AI.
Neuro-Inspired Vision-Language Models Show Resilience to Membership Inference Privacy Leakage
A new study explores whether neuro-inspired multi-modal vision-language models (VLMs) are resilient to membership inference privacy attacks. Using topological regularization, the authors found that NEURO VLMs reduce MIA success by up to 24% without sacrificing model utility, offering a promising path for secure AI deployment.
Neuro-Symbolic Framework Improves Motion Prediction for Autonomous Vehicles in Mixed Traffic
Researchers propose TraCS, a neuro-symbolic framework that augments black-box motion prediction with probabilistic first-order logic, improving accuracy and interpretability for autonomous vehicles in heterogeneous traffic. Tested on the Argoverse 2 benchmark, TraCS consistently improves state-of-the-art backbones.
Researchers Propose QoS-Aware Token Scheduling and Private Data Valuation for Multi-Modal Agentic Networks
A new arXiv paper introduces a QoS-aware token scheduling and private data valuation framework for decentralized multi-modal agentic networks. The approach embeds multi-modal data in a shared semantic space and uses differentially private prototypes to balance utility and privacy, showing improved fairness and QoS in simulations.