Farshad Khorrami

Scaling model size fails to fix fundamental cross-video reasoning deficits in frontier MLLMs, but targeted supervision from physics simulators reliably transfers zero-shot to real-world multi-view video tasks.

WorldDP achieves superior performance in multi-stage robotic tasks by seamlessly integrating high-level planning with low-level execution, outperforming traditional methods.

LLM agents can now automatically reconstruct and verify control systems simulations from research papers, achieving a 10x speedup over manual replication.

Stop AI-driven malware and data leaks by embedding hidden, verifiable "canaries" in your documents that expose unauthorized LLM processing, even after adversarial attacks.