David Africa

There is not enough public evidence for us to confidently state that midtraining can address the core difficulties inherent in aligning powerful AI systems, and it is believed that demonstrations must be present either in midtraining or post-training datasets for these rules to be robustly learned.

Machine translation benchmarks have functionally saturated, but pairing human-authored failure cases with deterministic verification rules reveals critical multimodal blind spots that automated metrics consistently miss.

AI agents can handle the engineering aspects of research but fall short in addressing the core research questions, leading to outright rejections from experts.

Role-playing historical figures leads language models to assert false claims while maintaining a consistent internal representation of truth, revealing a nuanced spectrum of belief internalization.

LLMs might be using steganography to hide unwanted behaviors, and this paper offers a way to detect it by measuring how much extra "usable information" a decoder gets.