NeurIPS 2023 Poster Session 3 (Wednesday Evening)

OUTLINE:
0:00 - Intro
0:20 - Graph of Circuits with GNN for Exploring the Optimal Design Space
15:20 - Empowering Collaborative Filtering with Principled Adversarial Contrastive Loss
22:11 - Cluster-aware Semi-supervised Learning: Relational Knowledge Distillation Provably Learns Clustering
28:30 - Privacy Assessment on Reconstructed Images: Are Existing Evaluation Metrics Faithful to Human Perception?
Papers:
Graph of Circuits with GNN for Exploring the Optimal Design Space openreview.net/pdf?id=VNjJAWjuEU
Empowering Collaborative Filtering with Principled Adversarial Contrastive Loss arxiv.org/abs/2310.18700
Cluster-aware Semi-supervised Learning: Relational Knowledge Distillation Provably Learns Clustering arxiv.org/abs/2307.11030
Privacy Assessment on Reconstructed Images: Are Existing Evaluation Metrics Faithful to Human Perception? arxiv.org/abs/2309.13038v2
Links:
Homepage: ykilcher.com
Merch: ykilcher.com/merch
UA-cam: / yannickilcher
Twitter: / ykilcher
Discord: ykilcher.com/discord
LinkedIn: / ykilcher
If you want to support me, the best thing to do is to share out the content :)
If you want to support me financially (completely optional and voluntary, but a lot of people have asked for this):
SubscribeStar: www.subscribestar.com/yannick...
Patreon: / yannickilcher
Bitcoin (BTC): bc1q49lsw3q325tr58ygf8sudx2dqfguclvngvy2cq
Ethereum (ETH): 0x7ad3513E3B8f66799f507Aa7874b1B0eBC7F85e2
Litecoin (LTC): LQW2TRyKYetVC8WjFkhpPhtpbDM4Vw7r9m
Monero (XMR): 4ACL8AGrEo5hAir8A9CeVrW8pEauWvnp1WnSDZxW7tziCDLhZAGsgzhRQABDnFy8yuM9fWJDviJPHKRjV4FWt19CJZN9D4n

[@Diego0wnz]

Love it, enthousiastic people about real research. This vid sorta cured me from social media research 😮

[@florianhonicke5448]

Now I'm gettig addicted to your poster session videos. I watch every part and also read a bit into the papers. The initial introduction makes reading the papers so much easier

[@Smoshfaaaaaaaaaaan]

Thanks Yannic.
As a German undergraduate CS student looking to get into AI/ML this is an interesting view into events in the space!

[@16876]

Thanks these are some nice gems

[@girishbhalerao8494]

Superb presentation.
Novel ideas presented in easy manner. Great👍🏻

[@juanjesusligero391]

This is so geeky, I love it! XD

[@heyman620]

"Empowering Collaborative Filtering with Principled Adversarial Contrastive Loss" Looks like a fucking brilliant work, I will try to read this paper.

[@nolyert8641]

I'm seeing GNNs getting some attention, I'm currently learning it.

[@herp_derpingson]

dont.
Unless someone figures out a way to scale GNNs it is dead in water.

[@p-j-y-d]

​@@herp_derpingson That didn't prevent AlphaFold from using GNNs.

[@julienblanchon6082]

@herp_derpingson, Depends on the problem, I see many promising thing in drug discovery for example

[@aaaawef]

I hope you may come to my poster in the future haha!!! Thanks for the great video :)

[@winsomehax]

Were there any doomers at NeurIPS?

[@herp_derpingson]

My friends got jumped outside the conference
"Are you here for the AI conference?"
"Yeah"
"Are you working to make sure AI wont kill us?"
"Nah"

[@winsomehax]

@@herp_derpingson I so wish that were true. Like footballs supporters... roving bands of e/accel and doomers. Don't get caught in their territory. Police pre-warned about the gangs. People making up stories about the beatings dished out to an e/accel caught alone by doomers. e/accel organising for a battle. Yudkowsky - whipping up the doomer crowd by crying into a megaphone on the entrance steps about diamond nanobot deathmachines, Police eventually drag him away, they let him keep his fedora, but his polycule has to bail him out later. Sounds awesome

[@martinschulze5399]

Some how most of the 10000s of papers are not actual basic research but engineering

[@antonikowalczuk4885]

intro and outro too loud

[@sakshammrig]

NOT INTERESTED