AI RESEARCH
Introducing Background Temperature to Characterise Hidden Randomness in Large Language Models
arXiv CS.AI
•
ArXi:2604.22411v1 Announce Type: new Even when decoding with temperature $T=0$, large language models (LLMs) can produce divergent outputs for identical inputs. Recent work by Thinking Machines Lab highlights implementation-level sources of nondeterminism, including batch-size variation, kernel non-invariance, and floating-point non-associativity. In this short note we formalize this behavior by