Episode Details

Today's papers: - Malicious Agent Skills in the Wild: A Large-Scale Security Empirical Study: https://arxiv.org/abs/2602.06547v1 - DAVE: Distribution-aware Attribution via ViT Gradient Decomposition: https://arxiv.org/abs/2602.06613v1 - TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering: https://arxiv.org/abs/2602.06911v1 - AIRS-Bench: a Suite of Tasks for Frontier AI Research Science Agents: https://arxiv.org/abs/2602.06855v3 - Toward generative machine learning for boosting ensembles of climate simulations: https://arxiv.org/abs/2602.06287v1 This podcast is from Colin Davis (colin-davis.com) using Claude & Elevenlabs.