The day in conversation for 2026-11-05
discussion
Transcript
The discussion: Ines: I think what really stands out from this batch is how MIST attempts to bypass the need for transcriptomics during inference by anchoring prototypes using those paired spatial data points; that cross-modal adaptation capability sounds powerful, but I’m curious how robust that molecular guidance actually remains when the frozen pathology model encounters completely novel cellular states it hasn't seen in training. Marcus: That robustness is tied directly to the feature space anchors MIST creates, which brings me to a point where I’m less sure about the agentic approach described in CTM-AI; if we rely on molecular priors from MIST, doesn't that inherently limit the agent's ability to discover fundamentally different biological mechanisms that don't conform to those initial expression anchors? For instance, how do you handle hypothesis generation when the molecular input space is artificially constrained by the training set’s transcriptomic map?
More episodes
- 2610.33881-Sensorimotor behavior from an untrained brain-and-nerve-cord connectome
- 2610.40344-Morphological Features of Murine Pulmonary Arterial Networks for Diagnosing Vascular Disease
- Daily Summary for 2026-11-19
- 2609.10193-Are You Learning Biological Signal or Shortcuts? Auditing and Mitigating Bias in Protein-Protein Interaction Datasets
- 2610.38236-Generative discovery of chemical reaction networks
- 2507.11176-An Interpretable AI framework Quantifying Traditional Chinese Medicine Principles Towards Enhancing and Integrating with Modern Biomedicine
- The day in conversation for 2026-11-19
- Daily Summary for 2026-11-19
- 2504.03590-Towards a Unified Framework for Determining Conformational Ensembles of Disordered Proteins
- 2603.15217-A multiscale discrete-to-continuum framework for structured population models