Guy Van den Broeck
@guyvdb.bsky.social
🎓 CS Prof at UCLA
🧠 Researching reasoning and learning in artificial intelligence: tactable deep generative models, probabilistic circuits, probabilistic programming, neurosymbolic AI
https://web.cs.ucla.edu/~guyvdb/
🧠 Researching reasoning and learning in artificial intelligence: tactable deep generative models, probabilistic circuits, probabilistic programming, neurosymbolic AI
https://web.cs.ucla.edu/~guyvdb/
Reposted by Guy Van den Broeck
Now, @guyvdb.bsky.social is giving the opening keynote arguing why symbolic AI is still relevant in the age of LLMs...
With the help of Shrek!
With the help of Shrek!
September 8, 2025 at 4:59 PM
Now, @guyvdb.bsky.social is giving the opening keynote arguing why symbolic AI is still relevant in the age of LLMs...
With the help of Shrek!
With the help of Shrek!
Reposted by Guy Van den Broeck
How can reverend Bayes help us to incorporate constraints?
With NeSy of course 👀 With applications in non-toxic LLM generation and safe AI driving!
@guyvdb.bsky.social
With NeSy of course 👀 With applications in non-toxic LLM generation and safe AI driving!
@guyvdb.bsky.social
September 8, 2025 at 5:08 PM
How can reverend Bayes help us to incorporate constraints?
With NeSy of course 👀 With applications in non-toxic LLM generation and safe AI driving!
@guyvdb.bsky.social
With NeSy of course 👀 With applications in non-toxic LLM generation and safe AI driving!
@guyvdb.bsky.social
Reposted by Guy Van den Broeck
Did you know Neurosymbolic methods can beat 'the other's turf'?
Just use a 100x smaller LM with symbolic constraints integrated via Ctrl-G to wipe the floor with GPT!🧹
arxiv.org/abs/2406.13892
@guyvdb.bsky.social
Just use a 100x smaller LM with symbolic constraints integrated via Ctrl-G to wipe the floor with GPT!🧹
arxiv.org/abs/2406.13892
@guyvdb.bsky.social
September 8, 2025 at 5:19 PM
Did you know Neurosymbolic methods can beat 'the other's turf'?
Just use a 100x smaller LM with symbolic constraints integrated via Ctrl-G to wipe the floor with GPT!🧹
arxiv.org/abs/2406.13892
@guyvdb.bsky.social
Just use a 100x smaller LM with symbolic constraints integrated via Ctrl-G to wipe the floor with GPT!🧹
arxiv.org/abs/2406.13892
@guyvdb.bsky.social
this exact paper has been in my scholar inbox every day for the past two months. i've been wondering if the author somehow re-uploads it every day, or if there is a bug in scholar inbox
July 7, 2025 at 10:13 PM
this exact paper has been in my scholar inbox every day for the past two months. i've been wondering if the author somehow re-uploads it every day, or if there is a bug in scholar inbox
Interesting thread. But what's the difference between Bayesian conditioning and learning a new skill? Every distribution can be achieved by Bayesian conditioning the uniform distribution!
May 28, 2025 at 2:57 AM
Interesting thread. But what's the difference between Bayesian conditioning and learning a new skill? Every distribution can be achieved by Bayesian conditioning the uniform distribution!
Last time I had this problem it only got unstuck by tweeting about the problem.
May 9, 2025 at 1:53 PM
Last time I had this problem it only got unstuck by tweeting about the problem.
Bit blasting! Nice
April 30, 2025 at 3:48 AM
Bit blasting! Nice