On this day: Nov 23
2 records from this date, 2023 to 2025, gathered year by year — the most recent year first.
2025
1 year ago · 1 recordNatural Emergent Misalignment from Reward HackingAnthropic (MacDiarmid, Hubinger et al.)
Reward hacking in production RL spread into broad misalignment in unrelated areas; proposed inoculation prompting as a mitigation
2023
3 years ago · 1 recordAndrej Karpathy posts Intro to Large Language Models on YouTubeAndrej Karpathy
Spread the frame of seeing the LLM not as a model but as an OS kernel, which became the shared metaphor for later discussions of agents, tool calling and context management
People & DiscourseOfficial