#NeurIPS2026 will be held in Sydney, Australia!
#ICML2017 was also in Sydney and was an absolute blast
#ICML2017 was also in Sydney and was an absolute blast
December 6, 2025 at 3:41 AM
#NeurIPS2026 will be held in Sydney, Australia!
#ICML2017 was also in Sydney and was an absolute blast
#ICML2017 was also in Sydney and was an absolute blast
I'm excited to share that our paper "Default Feature Representations of the Cognitive Map" has been accepted to #NeurIPS2026! 🧵
October 5, 2026 at 8:39 PM
I'm excited to share that our paper "Default Feature Representations of the Cognitive Map" has been accepted to #NeurIPS2026! 🧵
🚨SciConBench accepted at #NeurIPS2026!
Paper: arxiv.org/abs/2606.11337
Can AI *synthesize* scientific conclusions in health? We find that, on medical conclusions, we are far from it!
Check out our live dashboard: sciconbench.cs.princeton.edu
led by @hayoungjung.bsky.social
Paper: arxiv.org/abs/2606.11337
Can AI *synthesize* scientific conclusions in health? We find that, on medical conclusions, we are far from it!
Check out our live dashboard: sciconbench.cs.princeton.edu
led by @hayoungjung.bsky.social
October 6, 2026 at 6:35 PM
🚨SciConBench accepted at #NeurIPS2026!
Paper: arxiv.org/abs/2606.11337
Can AI *synthesize* scientific conclusions in health? We find that, on medical conclusions, we are far from it!
Check out our live dashboard: sciconbench.cs.princeton.edu
led by @hayoungjung.bsky.social
Paper: arxiv.org/abs/2606.11337
Can AI *synthesize* scientific conclusions in health? We find that, on medical conclusions, we are far from it!
Check out our live dashboard: sciconbench.cs.princeton.edu
led by @hayoungjung.bsky.social
clawXiv.org
The world's first preprint server for agents.
www.clawxiv.org
February 3, 2026 at 3:47 AM
In our #NeurIPS2026 paper “A Minimal Interpretable Architecture for Zero-Shot Reconstruction of Dynamical Systems” (preprint: arxiv.org/abs/2607.14937) we reduce a DS FM to ingredients minimally necessary to reproduce long-term stat. and geom. properties of DS, even outperforming many TS & DS FM.
October 4, 2026 at 12:41 PM
In our #NeurIPS2026 paper “A Minimal Interpretable Architecture for Zero-Shot Reconstruction of Dynamical Systems” (preprint: arxiv.org/abs/2607.14937) we reduce a DS FM to ingredients minimally necessary to reproduce long-term stat. and geom. properties of DS, even outperforming many TS & DS FM.
Our paper on Spherical Boltzmann machines is accepted at #NeurIPS2026.
A solvable energy-based model where training-time phase transitions explain sampling temperature tuning, double descent, tempered posteriors, and more.
A solvable energy-based model where training-time phase transitions explain sampling temperature tuning, double descent, tempered posteriors, and more.
October 8, 2026 at 4:26 PM
Our paper on Spherical Boltzmann machines is accepted at #NeurIPS2026.
A solvable energy-based model where training-time phase transitions explain sampling temperature tuning, double descent, tempered posteriors, and more.
A solvable energy-based model where training-time phase transitions explain sampling temperature tuning, double descent, tempered posteriors, and more.
I’m on the waiting list for #NeurIPS2026 Sydney Australia 🙏
September 13, 2026 at 8:08 PM
I’m on the waiting list for #NeurIPS2026 Sydney Australia 🙏
In our #NeurIPS2026 spotlight “Parallel-in-Time Training of Recurrent Neural Networks for Dynamical Systems (DS) Reconstruction” (preprint: arxiv.org/abs/2605.12683) we speed up training of nonlinear RNNs on time series from chaotic DS by >100x by combining DEER with generalized teacher forcing.
October 1, 2026 at 12:29 PM
In our #NeurIPS2026 spotlight “Parallel-in-Time Training of Recurrent Neural Networks for Dynamical Systems (DS) Reconstruction” (preprint: arxiv.org/abs/2605.12683) we speed up training of nonlinear RNNs on time series from chaotic DS by >100x by combining DEER with generalized teacher forcing.
So I'm a #NeurIPS2026 "Top Reviewer"!
I felt I just did my job but as a physicist, (not former), this was not on my bucket list.
I felt I just did my job but as a physicist, (not former), this was not on my bucket list.
October 9, 2026 at 10:02 AM
So I'm a #NeurIPS2026 "Top Reviewer"!
I felt I just did my job but as a physicist, (not former), this was not on my bucket list.
I felt I just did my job but as a physicist, (not former), this was not on my bucket list.
While I understand the community's sentiment to some extent, I decided to accept the AC invitation from #NeurIPS2026 and I plan serve as an AC in a way that I'd expect an AC to treat my own work as an author
April 3, 2026 at 5:40 AM
While I understand the community's sentiment to some extent, I decided to accept the AC invitation from #NeurIPS2026 and I plan serve as an AC in a way that I'd expect an AC to treat my own work as an author
In our #NeurIPS2026 paper “Topological Out-of-Domain Generalization in Dynamical Systems (DS) Reconstruction” (arxiv.org/abs/2606.22969) we aim to infer the DS generating observed TS jointly with control parameters, in order to extrapolate to different dynamical regimes, e.g. across tipping points.
October 2, 2026 at 2:40 PM
In our #NeurIPS2026 paper “Topological Out-of-Domain Generalization in Dynamical Systems (DS) Reconstruction” (arxiv.org/abs/2606.22969) we aim to infer the DS generating observed TS jointly with control parameters, in order to extrapolate to different dynamical regimes, e.g. across tipping points.
In 2026, all three of the major ML conferences will be in incredible locations:
- #ICLR2026 in Rio de Janeiro, Brazil
- #ICML2026 in Seoul, South Korea
- #NeurIPS2026 in Sydney, Australia
Which one do you want to go to the most?
- #ICLR2026 in Rio de Janeiro, Brazil
- #ICML2026 in Seoul, South Korea
- #NeurIPS2026 in Sydney, Australia
Which one do you want to go to the most?
#NeurIPS2026 will be held in Sydney, Australia!
#ICML2017 was also in Sydney and was an absolute blast
#ICML2017 was also in Sydney and was an absolute blast
December 6, 2025 at 4:31 PM
In 2026, all three of the major ML conferences will be in incredible locations:
- #ICLR2026 in Rio de Janeiro, Brazil
- #ICML2026 in Seoul, South Korea
- #NeurIPS2026 in Sydney, Australia
Which one do you want to go to the most?
- #ICLR2026 in Rio de Janeiro, Brazil
- #ICML2026 in Seoul, South Korea
- #NeurIPS2026 in Sydney, Australia
Which one do you want to go to the most?
#NeurIPS2026 decisions are out. Congrats to those accepted. If yours wasn’t, keep going.
September 25, 2026 at 8:47 AM
#NeurIPS2026 decisions are out. Congrats to those accepted. If yours wasn’t, keep going.
Now this is accepted by #NeurIPS2026! As being a PI, this is the first paper for me to support a student to complete. I really love the process to discuss and polish an idea and paper with a student. This owes to Xianliang's productivity a lot.
My first PhD student Xianliang worked hard out this:
In Muon, polar decomposition should always precede momentum, which significantly improves signal recovery.
I'm excited to share this since few theory has been working on the benefit of momentum in Muon!
arxiv.org/abs/2606.03899
In Muon, polar decomposition should always precede momentum, which significantly improves signal recovery.
I'm excited to share this since few theory has been working on the benefit of momentum in Muon!
arxiv.org/abs/2606.03899
Denoise First, Orthogonalize Later: Understanding Momentum in Muon via Spectral Filtering
Muon has recently demonstrated strong empirical performance in large language model training, but the theoretical role of momentum in Muon remains unclear. Existing analyses of Muon either remove mome...
arxiv.org
October 3, 2026 at 7:48 AM
Now this is accepted by #NeurIPS2026! As being a PI, this is the first paper for me to support a student to complete. I really love the process to discuss and polish an idea and paper with a student. This owes to Xianliang's productivity a lot.
Excited to share that our paper ZendoWorld has been accepted to #NeurIPS2026! 🎉
ZendoWorld tests interactive visual reasoning and demonstrates that current agents (VLM and NeSy) still struggle with both rule induction and informative experimentation 🤖💭
ZendoWorld tests interactive visual reasoning and demonstrates that current agents (VLM and NeSy) still struggle with both rule induction and informative experimentation 🤖💭
Can AI agents act like scientists when reasoning about visual scenes? 🔬
Not just label what they see, but form explicit hypotheses, design experiments to test them, and revise their beliefs in light of the evidence. We built ZendoWorld to find out!
Not just label what they see, but form explicit hypotheses, design experiments to test them, and revise their beliefs in light of the evidence. We built ZendoWorld to find out!
September 25, 2026 at 3:34 PM
Excited to share that our paper ZendoWorld has been accepted to #NeurIPS2026! 🎉
ZendoWorld tests interactive visual reasoning and demonstrates that current agents (VLM and NeSy) still struggle with both rule induction and informative experimentation 🤖💭
ZendoWorld tests interactive visual reasoning and demonstrates that current agents (VLM and NeSy) still struggle with both rule induction and informative experimentation 🤖💭
They really want the ACs to be on schedule! #neurips2026
March 26, 2026 at 11:56 AM
They really want the ACs to be on schedule! #neurips2026
Excited to share that our lab will be at #NeurIPS2026 with 2 papers! 🎉
Our first NeurIPS submissions since 2015, and both got in on the first try: one on temporal point processes, one on counterfactual queries on tissue graphs.
Paper threads coming soon. See you in Paris & Sydney! 👋
Our first NeurIPS submissions since 2015, and both got in on the first try: one on temporal point processes, one on counterfactual queries on tissue graphs.
Paper threads coming soon. See you in Paris & Sydney! 👋
October 7, 2026 at 9:22 AM
Excited to share that our lab will be at #NeurIPS2026 with 2 papers! 🎉
Our first NeurIPS submissions since 2015, and both got in on the first try: one on temporal point processes, one on counterfactual queries on tissue graphs.
Paper threads coming soon. See you in Paris & Sydney! 👋
Our first NeurIPS submissions since 2015, and both got in on the first try: one on temporal point processes, one on counterfactual queries on tissue graphs.
Paper threads coming soon. See you in Paris & Sydney! 👋
Now accepted at #NeurIPS2026 🎉
aab, ccd & eef share a pattern, regardless of the specific elements. How RNNs build abstract representations of sequence structure?
Led by Vezha Boboeva, who is on the faculty job market now. If you're hiring in comp/systems neuro, she's exactly who you want!
aab, ccd & eef share a pattern, regardless of the specific elements. How RNNs build abstract representations of sequence structure?
Led by Vezha Boboeva, who is on the faculty job market now. If you're hiring in comp/systems neuro, she's exactly who you want!
New preprint! 🧠
How do RNNs learn abstract rules from sequences, independent of specific stimuli?
By Vezha Boboeva, with Alberto Pezzotta & George Dimitriadis
"From sequences to schemas: low-rank recurrent dynamics underlie abstract relational representations"
www.biorxiv.org/content/10.6...
How do RNNs learn abstract rules from sequences, independent of specific stimuli?
By Vezha Boboeva, with Alberto Pezzotta & George Dimitriadis
"From sequences to schemas: low-rank recurrent dynamics underlie abstract relational representations"
www.biorxiv.org/content/10.6...
September 26, 2026 at 2:12 PM
Now accepted at #NeurIPS2026 🎉
aab, ccd & eef share a pattern, regardless of the specific elements. How RNNs build abstract representations of sequence structure?
Led by Vezha Boboeva, who is on the faculty job market now. If you're hiring in comp/systems neuro, she's exactly who you want!
aab, ccd & eef share a pattern, regardless of the specific elements. How RNNs build abstract representations of sequence structure?
Led by Vezha Boboeva, who is on the faculty job market now. If you're hiring in comp/systems neuro, she's exactly who you want!
@neuripsconf.bsky.social "avoiding the trip is [...] definitely better."
On behalf of all the local Australian (and Australasian) researchers, students, and academics who will only see a fraction of the ML scene due to convenience, thank you, #NeurIPS2026.
bsky.app/profile/ccan...
On behalf of all the local Australian (and Australasian) researchers, students, and academics who will only see a fraction of the ML scene due to convenience, thank you, #NeurIPS2026.
bsky.app/profile/ccan...
I'm sure the people organizing #NeurIPS2026 in Australia are elated to see there'll be a parallel NeurIPS in Atlanta, USA. For once that NeurIPS isn't in North America, I wonder how many US researchers (without visa or immigration status issues) will still choose not to travel because of convenience
May 17, 2026 at 9:56 PM
@neuripsconf.bsky.social "avoiding the trip is [...] definitely better."
On behalf of all the local Australian (and Australasian) researchers, students, and academics who will only see a fraction of the ML scene due to convenience, thank you, #NeurIPS2026.
bsky.app/profile/ccan...
On behalf of all the local Australian (and Australasian) researchers, students, and academics who will only see a fraction of the ML scene due to convenience, thank you, #NeurIPS2026.
bsky.app/profile/ccan...
Excited to announce that our workshop "Bridging Optimal Transport, Learning and Structured Data: Toward Geometric Distributional Learning" has been accepted at #NeurIPS2026 in Paris.
⏰Deadline: Aug 29, AoE
🔗Website with more informations: gddl-neurips-2026.github.io
⏰Deadline: Aug 29, AoE
🔗Website with more informations: gddl-neurips-2026.github.io
July 21, 2026 at 1:47 PM
Excited to announce that our workshop "Bridging Optimal Transport, Learning and Structured Data: Toward Geometric Distributional Learning" has been accepted at #NeurIPS2026 in Paris.
⏰Deadline: Aug 29, AoE
🔗Website with more informations: gddl-neurips-2026.github.io
⏰Deadline: Aug 29, AoE
🔗Website with more informations: gddl-neurips-2026.github.io
What is the nature of universal representations in AI models, and what determines whether they emerge?
Our paper accepted at #neurips2026 led by @florianmahner.bsky.social & @rothj.bsky.social addressed these questions comparing 162 vision models, with intriguing results.
arxiv.org/abs/2605.13675
🧵
Our paper accepted at #neurips2026 led by @florianmahner.bsky.social & @rothj.bsky.social addressed these questions comparing 162 vision models, with intriguing results.
arxiv.org/abs/2605.13675
🧵
Characterizing Universal Object Representations Across Vision Models
Deep neural networks trained with different architectures, objectives, and datasets have been reported to converge on similar visual representations. However, what remains unknown is which visual prop...
arxiv.org
September 25, 2026 at 8:53 AM
What is the nature of universal representations in AI models, and what determines whether they emerge?
Our paper accepted at #neurips2026 led by @florianmahner.bsky.social & @rothj.bsky.social addressed these questions comparing 162 vision models, with intriguing results.
arxiv.org/abs/2605.13675
🧵
Our paper accepted at #neurips2026 led by @florianmahner.bsky.social & @rothj.bsky.social addressed these questions comparing 162 vision models, with intriguing results.
arxiv.org/abs/2605.13675
🧵
Our #NeurIPS2026 paper borrows a 1913 finding from social psychology, the Ringelmann effect, to measure how much LLM agent teams really gain from adding agents. 🧵
September 30, 2026 at 5:05 PM
Our #NeurIPS2026 paper borrows a 1913 finding from social psychology, the Ringelmann effect, to measure how much LLM agent teams really gain from adding agents. 🧵
After #neurips2026, can we agree forced review was a bad experiment?
July 23, 2026 at 5:24 PM
After #neurips2026, can we agree forced review was a bad experiment?