Uploaded October 2026 | Updated November 2026, 2 days ago
Grigoris Velegkas (Google Research)
https://simons.berkeley.edu/talks/grigoris-velegkas-google-research-2026-10-05
Trustworthy AI: From Hallucinations to Reliable Autonomy
Modern language models are expected to generate outputs that are both valid and diverse, even though they are trained on imperfect data. We study these goals through the framework of language generation in the limit. We first ask whether a generator trained on correct data can cover the full breadth of its target language while avoiding hallucinations. We show that these goals conflict in general: requiring only valid outputs can force the generator to collapse onto a narrow part of the language. We then study generation from data containing incorrect examples and characterize how the density of these errors affects diverse and reliable generation. Finally, since hallucinations may be unavoidable in certain settings, we ask whether they can at least be detected. We show that this problem is tightly connected to the classical setting of language identification in the limit.
This talk is based on joint work with Alkis Kalavasis, Amin Karbasi, Anay Mehrotra, Omar Montasser, John Sous, Xifan Yu, and Felix Zhou.
Grigoris Velegkas (Google Research)
https://simons.berkeley.edu/talks/grigoris-velegkas-google-research-2026-10-05
Trustworthy AI: From Hallucinations to Reliable Autonomy
Modern language models are expected to generate outputs that are both valid and diverse, even though they are trained on imperfect data. We study these goals through the framework of language generation in the limit. We first ask whether a generator trained on correct data can cover the full breadth of its target language while avoiding hallucinations. We show that these goals conflict in general: requiring only valid outputs can force the generator to collapse onto a narrow part of the language. We then study generation from data containing incorrect examples and characterize how the density of these errors affects diverse and reliable generation. Finally, since hallucinations may be unavoidable in certain settings, we ask whether they can at least be detected. We show that this problem is tightly connected to the classical setting of language identification in the limit.
This talk is based on joint work with Alkis Kalavasis, Amin Karbasi, Anay Mehrotra, Omar Montasser, John Sous, Xifan Yu, and Felix Zhou.










