N. Kriegeskorte

Computation time distributions, not just scaling, reveal the underlying mechanisms of physical computation across diverse systems.

Models trained with VBVR-Pro not only excel in a controlled task space but also show significant transferability to external benchmarks, revealing critical insights into visual reasoning mechanisms.

LLMs encode truthfulness along a spectrum from general principles to narrow domain expertise, and exploiting domain-specific truth representations is key to steering model behavior.