Probing VLM Medical Image Recognition with Silico Platform
mathildepapillo · x · 2026-08-05
Using Silico, an AI interpretability platform by Goodfire, a developer trained a J-Lens and built a visualizer to explore what the Qwen-3.5 27B model focuses on across different patches of medical images. The experiment revealed that the model could successfully identify details like a harmful cyst on an OCT scan.
More from Research
- Goodfire Launches Silico: AI Agents Can Autonomously Run Experiments for Days — mathildepapillo · 2026-08-05
- Impressive Progress on NIST Robot Benchmark: Models Master Complex Contact-Rich Tasks — chris_j_paxton · 2026-08-05
- Visualizing Qwen2.5-VL's "Thoughts" Using Goodfire Silico — ninamiolane · 2026-08-05
- Google's AI Overview Flips Answer Based on Single arXiv Preprint — sayashk · 2026-08-05
- Cursor Releases Mixture-of-Kittens Megakernel for MoE, Claims Nearly 2x TFLOP/s — CapnHat · 2026-08-05
- OpenADMET Launches 3rd Challenge: Predicting CYP Inhibition — CatAstro_Piyush · 2026-08-05