3 ms·
This is all obvious and not news, but there's a lot of people doing fighting retreats against LLM intelligence where the degree of obviousness matters.
by FeepingCreature 3mo ago
This is all obvious and not news, but there's a lot of people doing fighting retreats against LLM intelligence where the degree of obviousness matters.
- XenophileJKO 3mo agoAnticipatory token selection was obvious when you used structured output on gpt 3.5. The model would construct/hallucinate pre-conditions to satisfy some final output value that the model was predisposed to.