Uploaded October 2026 | Updated November 2026, 4 days ago
Santosh Vempala (Georgia Institute of Technology)
https://simons.berkeley.edu/talks/santosh-vempala-georgia-institute-technology-2026-10-07
Trustworthy AI: From Hallucinations to Reliable Autonomy
We analyze hallucination in language models from a mathematical perspective and find that the statistical pressures of the next-word prediction training pipeline induce hallucinations under fairly general conditions even when trained on clean data. The core proof is a simple reduction from a natural classification task ("fact or hallucination") to a generation/prompt completion task.
The talk is based on facts and on work with Adam T. Kalai (STOC 2025) and with Kalai, Ofer Nachum and Eddie Zhang (Nature 2026).
Santosh Vempala (Georgia Institute of Technology)
https://simons.berkeley.edu/talks/santosh-vempala-georgia-institute-technology-2026-10-07
Trustworthy AI: From Hallucinations to Reliable Autonomy
We analyze hallucination in language models from a mathematical perspective and find that the statistical pressures of the next-word prediction training pipeline induce hallucinations under fairly general conditions even when trained on clean data. The core proof is a simple reduction from a natural classification task ("fact or hallucination") to a generation/prompt completion task.
The talk is based on facts and on work with Adam T. Kalai (STOC 2025) and with Kalai, Ofer Nachum and Eddie Zhang (Nature 2026).










