Sebastian Prasanna

An Em­piri­cal Study of Meth­ods for SFTing Opaque Rea­son­ing Models

How do LLMs gen­er­al­ize when we do train­ing that is in­tu­itively com­pat­i­ble with two off-dis­tri­bu­tion be­hav­iors?

Five ap­proaches to eval­u­at­ing train­ing-based con­trol measures

Model or­ganisms re­searchers should check whether high LRs defeat their model organisms

Ba­sic Leg­i­bil­ity Pro­to­cols Im­prove Trusted Monitoring