Blog

This is a space to share technical research updates, blog about life, or simply collect any half-formed thoughts that I feel compelled to write down. I think putting things here is in of itself helpful (even if nobody actually reads it), simply because writing is also an epistemological process. Writing here helps me take my unstructured ideas and (gradually) transform them into a better, more structured understanding. I might occasionally crosspost these things to other forums.

Constitutional Judges Disprefer Evaluation Aware Transcripts
Investigating the impact of CoT leakage
Probes for Evaluation Awareness
Ant Morality
Updating Toward Legibility: A Calibrated, Tractable, Neglected Investigation Into the Upstream Diffusion of Load-Bearing Rationalist Jargon, Operationalized