A non-sofic group @RyanODonnellTeaching
A non-sofic group  @RyanODonnellTeaching
Uploaded September 2026 | Updated November 2026, 3 hours ago
A talk at the Carnegie Mellon ACO seminar in the math department about the existence of a non-sofic group, proven by OpenAI (based in part by work of Kun and Thom).

cdn.openai.com/pdf/ten-proofs-oai.pdf
arxiv.org/pdf/2608.06222

Link to the note I wrote about Property (T): https://www.cs.cmu.edu/~odonnell/papers/elementary-property-t.pdf

Apologies for any and all mistakes. Also, the proof sketch I presented is not identical to any of the published proofs, but is a tiny massaging of them.
A non-sofic group#33/100: Hadamard transform: how it works || Quantum Computer Programming in 100 Easy LessonsThe Central Binomial Coefficient || @ CMU || Recitation 1 of CS Theory Toolkit#50/100: Discriminating two rotations || Quantum Computer Programming in 100 Easy Lessons#11/100: Amplitude trees || Quantum Computer Programming in 100 Easy Lessons#61/100: Detecting medium 1-qubit rotations || Quantum Computer Programming in 100 Easy Lessons#35/100: Kets || Quantum Computer Programming in 100 Easy Lessons#47/100: Discriminating 2 qubits: optimal error || Quantum Computer Programming in 100 Easy Lessons#31/100: Bitmasked XOR functions || Quantum Computer Programming in 100 Easy Lessons#92/100: Recapping Factoring technical details || Quantum Computer Programming in 100 Easy Lessons#7/100: Classical reversible instructions || Quantum Computer Programming in 100 Easy LessonsFiber bundle codes: Breaking the N¹ᐟ² polylog(N) barrier for quantum LDPC codes
Ryan ODonnell |

A non-sofic group

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER