@benh @futurebird
@chris_evelyn
Brainstorming with the help of an LLM has been tested, the results are worse than without
@benh @futurebird
@chris_evelyn
Brainstorming with the help of an LLM has been tested, the results are worse than without
@futurebird @bri_seven
Exactly — they can bolt a thing to the output part of an LLM to force it to only output valid python programs, but it doesn't make the LLM any smarter; it just forces it to output valid python programs
@bri_seven @futurebird
Presumably constraining the LLM output to valid python programs at the sampling stage, potentially giving a misleading impression about its capability