Greedy decoding in Large Language Models, which selects the argmax token at each step, produces locally optimal but globally inconsistent outputs and is prone to repetition or loops.

Authors

• Person: M. Brenndoerfer Organization: mbrenndoerfer.com
  Hallucination Causes: Why Language Models Fabricate Facts

Sources

• Hallucination Causes: Why Language Models Fabricate Facts mbrenndoerfer.com M. Brenndoerfer · mbrenndoerfer.com via serper

Referenced by nodes (1)

• Large Language Models concept