AI Papers - 2026-10-28

DailyArxiv - AI Research Podcast · Oct 28, 2026 · 14m

Today's papers: - Log-To-Leak: Prompt Injection Attacks on Tool-Using LLM Agents via the Model Context Protocol: https://arxiv.org/abs/2610.36108v1 - Patient-specific rubric-based evaluation of large language models across longitudinal cancer care: https://arxiv.org/abs/2610.35847v1 - Timestral: https://arxiv.org/abs/2610.37585v1 - Pretraining Knows What, Simulation Teaches How: Compositional Generalization in a Sim-Trained Robot Agent: https://arxiv.org/abs/2610.35817v2 - VideoLVR: Latent Visual Reasoning over Grounded Video Evidence: https://arxiv.org/abs/2610.37613v1 This podcast is from…