On Friday, 2 January 2026, at 1:23 AM, Matt Mahoney wrote: > What happens if you tell the LLM to ignore the token flags? How could you > test an LLM to make sure this can't happen? Have you done any actual tests? I haven't tested anything yet, just chatted with Claude about the idea...
> What happens if you tell the LLM to ignore the token flags? Within the "green" tokens (data)? Nothing should happen. That's trained as part of the adversarial examples in the command following dataset. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T2faee51273b20a92-M8a8756cc8483648160df98ca Delivery options: https://agi.topicbox.com/groups/agi/subscription
