Llama-3.1-8B-Instruct

Product · 1 documents

Key number

96%

Identity share of advice variance at zero disclosure. arXiv (Cornell University) p.1

Thin Evidence, Thick Priors: How Language Models Substitute Identity for Missing Financial Facts

Research paperarXiv (Cornell University)October 202649 pages

Across 96,600 prompts to Llama-3.1-8B-Instruct, the average identity swing in recommended equity allocation rises from 4.78 percentage points at full disclosure to 10.34 percentage points when no financial facts are supplied.