Paper Tests LLM Probabilistic Coherence via Stock Arbitrage

A new paper by Suproteem Sarkar and economist Isaiah Andrews measures LLM probabilistic coherence using historical stock returns, quantifying arbitrage profits from inconsistent predictions. It finds that more logical dependencies between events lead to less coherent model predictions.

2026-09-05 ~ 2026-09-05 · 2 related posts

• Measuring LLM Forecast Incoherence via Arbitrage Profits: New Paper from Sarkar & Andrews — soumitrashukla9 · 2026-09-05
• New Paper: LLM Forecasts Get More Incoherent as Logical Relations Between Events Increase — soumitrashukla9 · 2026-09-05