sschmitt

↗ HN profile [ 72.8 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. Do you like my definition of consciousness? while True: context += LLM(context) Using the idea of consciousness as the interface to oneself as proposed by Daniel Dennett.

  2. context = "you are a useful model. you evaluate your own output for N steps after the initial user input" prompt = "here comes a prompt by the user" context += prompt for _ in range(N): context += evaluate_llm(context)

  3. GPT-4 supports a 128,000-token context window (per OpenAI). I'm trying to understand how much text that really is. Here's my rough estimate: 1 token ≈ ¾ word ≈ 4 characters A single-spaced A4 page (12pt font) ≈ 500–600…

  4. From clean room to machine room (iopscience.iop.org)