← Dictionary / Sampling Temperature

Sampling Temperature

Appears in 1 paper

A hyperparameter in language model decoding that controls randomness.

As used in Paper 23 — Scaling LLM Test-Time Compute Optimally Can be More Effective than Scaling Model Parameters →

A hyperparameter in language model decoding that controls randomness. High temperature (e.g., 1.0) produces more diverse, creative generations. Low temperature (e.g., 0.1) produces more deterministic, conservative outputs. Best-of-N typically uses higher temperature to ensure diversity across samples.

Appears in papers

Paper 23 — Scaling LLM Test-Time Compute Optimally Can be More Effective than Scaling Model Parameters →