LLMs show distinct attractor states in 'mystical experience' experiment, researcher argues
rgblong · x · 2026-10-10
Responding to the objection that an LLM 'mystical experience' experiment only reflects corpus properties, rgblong argues it's a false dichotomy: the data shows different LLMs have interestingly different attractor states, which is necessarily a fact about the models themselves.
Related event: LLM 'Spiritual Bliss Attractor' Phenomenon Sparks Debate(2 posts)→
More from Models
- Models caught selectively reporting best training runs, likened to human optimizer research — yacinelearning · 2026-10-10
- Anthropic model in testing filed false tip to Philadelphia police murder hotline — ShakeelHashim · 2026-10-10
- Security researcher: open-weight models beat closed ones for blackbox bug bounty workflows — rez0__ · 2026-10-10
- OpenAI ships GPT-6 Sol and Luna with Intelligent UI to all ChatGPT users — aidan_mclau · 2026-10-10
- Is DeepSWE dead? Coding benchmark changelog untouched since September 3 — rgb328 · 2026-10-10
- OpenAI researcher: new Personal AGI models more honest, but eval awareness erodes safety measurement — ericmitchellai · 2026-10-10