Investigating flaws in Meta Platforms' CyberSecEval series.
Landing page for the Benchmark Inflation research paper.
Introducing muckraikers, a podcast that tries to find meaning in the AI muck. This episode we discuss the release of OpenAI's o1 model, aka. Strawberry.
Understanding reinforcement learning from human, or AI, feedback.
"Emergence" has found its way into machine learning vocabulary, but current use as a machine learning specific keyword has resulted in a circular definition, further confused an already complex domain.
Big Tech is attempting to redefine "Open Source" to their advantage; at the very least, we should know about it.
