3 ms·
To me that makes it worse. Why would two people get wildly different answers to a simple factual observation query.
by JasonBee 1y ago
To me that makes it worse. Why would two people get wildly different answers to a simple factual observation query.
- nostrebored 1y agoBecause of the interplay of how tokenizers work, temperature, and adaptive reasoning? These models aren't fact generators.
- vaenaes 1y ago~stochasticity~