The ‘truth serum’ for AI: OpenAI’s new method for training models to confess their mistakes
FeaturedBen Dickson December 4, 2025 OpenAI researchers have introduced a novel method that acts as a “truth serum” for large…
training
MIT offshoot Liquid AI releases blueprint for enterprise-grade small-model training
Carl Franzen December 1, 2025 Credit: VentureBeat made with Google Nano Banana Pro using FAL.aiWhen Liquid AI, a startup founded…
Researchers show that training on “junk data” can lead to LLM “brain rot”
Text settings Story text Size Small Standard Large Width * Standard Wide Links Standard Orange * Subscribers only Learn more…
A small number of samples can poison LLMs of any size
AlignmentA small number of samples can poison LLMs of any sizeOct 9, 2025Read the paperIn a joint study with the…
Anthropic users face a new choice – opt out or share your data for AI training
Anthropic is making some big changes to how it handles user data, requiring all Claude users to decide by September…