AI Papers - 2026-04-08

Today's papers: - StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing: https://arxiv.org/abs/2604.05014v1 - QED-Nano: Teaching a Tiny Model to Prove Hard Theorems: https://arxiv.org/abs/2604.04898v1 - Thinking Diffusion: Penalize and Guide Visual-Grounded Reasoning in Diffusion Multimodal Language Models: https://arxiv.org/abs/2604.05497v1 - MedGemma 1.5 Technical Report: https://arxiv.org/abs/2604.05081v1 - One Model for All: Multi-Objective Controllable Language Models: https://arxiv.org/abs/2604.04497v1 This podcast is from Colin Davis (colin-davis.com) using Claude & Elevenlabs.