BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Jordan Ellenberg (University of Wisconsin)
DTSTART:20240206T210000Z
DTEND:20240206T220000Z
DTSTAMP:20260423T125532Z
UID:MITNT/85
DESCRIPTION:Title: <a href="https://researchseminars.org/talk/MITNT/85/">A
 round Smyth’s conjecture</a>\nby Jordan Ellenberg (University of Wiscons
 in) as part of MIT number theory seminar\n\nLecture held in Room 2-449 in 
 the Simons Building (building 2).\n\nAbstract\nAre there algebraic numbers
  $x\,y\,z$ which are Galois conjugate to each other over $\\mathbb{Q}$ and
  which satisfy the equation $5x + 6y + 7z = 0$? In 1986\, Chris Smyth prop
 osed an appealingly simple conjecture about linear relations between Galoi
 s conjugates\, which would provide answers to the above questions and all 
 questions of the same form\, and which has remained unsolved.  My experien
 ce is that most people\, upon seeing Smyth’s conjecture\, immediately th
 ink it must be false (I certainly did!)\, but I have come to think it’s 
 true\, and I’ll talk about a provisional solution\, joint with Will Hard
 t\, in the case of three conjugates.  I’ll explain why (as Smyth observe
 d) this is really a conjecture about linear combinations of permutation ma
 trices (related question\, solved by Speyer:  which algebraic numbers can 
 be eigenvalues of the sum of two permutation matrices?)\, and why our appr
 oach can be thought of as proving a “Hasse principle for probability dis
 tributions” in a particular case\, plus a bit of additive number theory.
   Much of this talk\, maybe all\, will be suitable for undergraduates.\n
LOCATION:https://researchseminars.org/talk/MITNT/85/
END:VEVENT
END:VCALENDAR
