DisasterTD: Disaster Toponym Disambiguation Using Multimodal LLMs and Cross-View Geolocalization CoTinyVLA: Chain-of-Thought Distillation for a Sub-Billion-Parameter Vision-Language-Action Model Inferring Missing Trajectory Data with Temporal Convolutional Networks Toward Standardized Cross-Vendor Agent Tool Trust Management in Autonomous Networks Evaluating Communicative Belief Updates in Large Language Models via Implicature Recognition and Cancellation