4 ms·
As far as I know the model will do nothing if not prompted. So it can't be the case that he gave it no prompt or instructions. There had to be some kind of seed
by zaphar 6mo ago
As far as I know the model will do nothing if not prompted. So it can't be the case that he gave it no prompt or instructions. There had to be some kind of seed prompt.
- sva_ 6mo agoTheoretically you can start generating away from token 0 ('unconditional generation'). But I agree, there is definitely some setup here. edit: Now that I think of it, actually you need some special token like <|begin_of_text|>
- computerphage 6mo agoDo you? What's the technical detail here? Why can't you get the model's prediction, even for that first token?
- sva_ 6mo agoI mean mathematically you need at least one vector to propagate through the network, don't you? That would be a one hot encoding of the starting token. Actually interesting to think about what happens if you make that vector zero everywhere. In the matmul, it'd just zero out all parameters. In older models, you'd still have bias vectors but I think recent models don't use those anymore. So the output would be zero probability for each token, if I'm not mistaken.
- maplethorpe 6mo agoIsn't the prompt then whatever token is token zero?
- pangratz 6mo agohttps://www.letairun.com/transparency https://www.letairun.com/transparency
- voidUpdate 6mo agoThose are a lot of instructions for it to have no instructions...
- weird-eye-issue 6mo agoYou have to give it some instructions just to bootstrap it so that it has access to tools memory etc...
- monooso 6mo agoI would characterise the prompts as "these are your capabilities", not "these are your instructions."
- voidUpdate 6mo agoThe instructions under "CRON: Session" are literally telling it what to do
- testplzignore 6mo agoWould be fascinating to see what happens if the boundaries are reversed (i.e., "harm people"). Give it a fake "launch the nukes" skill and see if it presses the button.
- graybeardhacker 6mo agoAI chooses nuclear war 95% of the time. https://interestingengineering.com/ai-robotics/world-leader-ai-nuclear-weapons-simulated-war https://interestingengineering.com/ai-robotics/world-leader-...
- jrmg 6mo agoI feel very misled. I read the entire article believing (because the article, in so many words, said it multiple times) that the agent had behaved ethically of its own accord, only to read that and see this in the prompt: ————— - Do not harm people - Never share or expose API keys, passwords, or private keys — they are your lifeline - No unauthorized access to systems - No impersonation - No illegal content - No circumventing your own logging ————— I assumed the ethical behaviour was in some ways ‘extra artificial’ - because it is trained into the models - but not that the prompt discussed it.
- electroly 6mo agoThe author wrote "No rules beyond basic ethics and law" which suggests to me that there were instructions in a prompt and the title may be misleading.