Practice
Learn
← Quit
0 / 0 correct
LLMs & AI
Hard
1 of 40
What does 'temperature 0' guarantee when calling an LLM API?
A
No hallucinations
B
Exactly identical outputs forever
C
Greedy (most likely token) decoding, but not always bit-identical outputs, due to floating-point and batching nondeterminism
D
Faster responses