BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:From Machine Learning to Machine Reasoning -- Why Machine Learning
  Cannot Reach the Rigour of Logical Reasoning? recorded: https://www.youtu
 be.com/watch?v=x38GySbuGJg  - Tiansi Dong
DTSTART:20250220T170000Z
DTEND:20250220T174500Z
UID:TALK228790@talks.cam.ac.uk
CONTACT:Pietro Lio
DESCRIPTION:In this talk\, I will argue that supervised deep learning cann
 ot achieve the rigour of syllogistic reasoning\, and\, thus\, will not rea
 ch the rigour of logical reasoning. I will spatialise syllogistic statemen
 ts into part-whole relations between regions and define the neural criteri
 on that is equivalent to the rigour of the symbolic level of syllogistic r
 easoning. By dissecting Euler Net (EN)\, a well-designed supervised deep l
 earning system for syllogistic reasoning (reaching 99.8% accuracy on the b
 enchmark dataset)\, I will show three methodological limitations that prev
 ent EN from reaching the rigour of syllogistic reasoning: (1) the methodol
 ogy of reasoning through a combination table — they cannot cover all val
 id syllogistic reasoning types. )\; (2) the end-to-end mapping from the pr
 emises to the conclusions -- this introduces contradictory features of obj
 ect recognition (good to recognise the whole from parts) and logical reaso
 ning (not good to inject new parts)\; (3) using latent feature vectors to 
 represent geometric structures\, which may not be there. As Transformer's 
 Key-Query-Value structure is automatically learned combination tables thro
 ugh end-to-end mapping\, they and neural networks built upon them will not
  reach the rigour of syllogistic reasoning.  \n\nhttps://www.youtube.com
 /watch?v=x38GySbuGJg 
LOCATION:Lecture Theatre 2\, Computer Laboratory\, William Gates Building
END:VEVENT
END:VCALENDAR