One advantage of holding NeurIPS conference this year in San Diego as opposed to Vancouver is that San Diego isn't freezing cold.

Memories of #NeurIPS2024!

🎧 Hear Wanru Zhao talk CLUES on WiAIR – Women in AI Research:
🎥 YouTube: www.youtube.com/watch?v=pq4k...
🎙️ Spotify: open.spotify.com/episode/3qWk...
🎧 Apple: podcasts.apple.com/ca/podcast/d...
#WiAIR #NeurIPS2024 #LLMs #FederatedLearning #CLUES #WomenInAI #AIResearch

If I remember correctly, that was also the first CV conference with over 1000 papers, and people already felt overwhelmed. Now, CVPR 2025 has 2800+ papers, and #NeurIPS2024 had 4497. It’s becoming nearly impossible to discover hidden gems while wandering poster sessions. 2/2

📝 Peer reviewing is often an invisible part of research. It's 🧠-demanding and ⏳-consuming, but essential to maintain/improve the quality/rigor of our fields.

⭐️🎉 Glad to be recognized as a Top Reviewer at #NeurIPS2024 and Outstanding Reviewer at #CVPR2025

🌍 How do #LLMs handle trolley problems across cultures? We test them with 98K dilemmas in 107 languages, grounded in 40M+ human moral judgments.
💡 Spotlight @ICLR2025 (in Singapore✈️) | Best Paper #NeurIPS2024 Pluralistic Alignment Workshop
📄 Check out our paper: arxiv.org/abs/2407.02273 🧵👇

Check out our very popular "Tutorial on Causality for LLMs" at #NeurIPS2024: 2-hour detailed intro + a super insightful panel w/ @Yoshua_Bengio @bschoelkopf @Swarooprm7 @giambattista92 @tom4everitt on "The Role of Causality in Future LLMs" 📺 youtube.com/watch?v=8j1f... with slides

Not sure ICML/NeurIPS/ICLR are appropriate venues for all CV papers. Although lately it appears more vision papers are finding traction in these conferences. My take CV researchers are looking for more opportunities to publish. Check out this plot from #NeurIPS2024.

Releasing EgoSim, a simulator for body-worn cameras. #NeurIPS2024

EgoSim takes real mocap data (e.g., from AMASS) and synthesizes multi-modal egocentric videos.

Plus: MultiEgoView, a real dataset from 6 GoPro cameras and ground-truth 3D poses during several activities (13 people).

Pasan's research focuses on reconstructing simpler, interpretable, and efficient models from larger complex models by specifically integrating explainability (recent publications at #NeurIPS2024 #AISTATS2025 #AAMAS2025)

🚨 Image AutoRegressive Models Leak More Training Data Than Diffusion Models🚨

IARs — like the #NeurIPS2024 Best Paper — now lead in AI image generation. But at what risk?

IARs:
🔍 Are more likely than DMs to reveal training data
🖼️ Leak entire training images verbatim

🧵 1/

Two researchers from the Hertie AI Data Science department presented their posters at #NeurIPS2024 in Vancouver.

🔍 Recap of #NeurIPS2024 in Vancouver:

Hertie AI members Ifeoma Nwabufo and @sdamrich.bsky.social presented their work at the largest @neuripsconf.bsky.social conference to date.

Read the full article here: hertie.ai/two-research...

#MachineLearning #AIResearch #HertieAI #WIMLDS

#NeurIPS2024

Interview with Yuki Mitsufuji: Improving AI image generation - ΑΙhub

Great to chat to Yuki Mitsufuji (Sony AI) about two pieces of research presented at #NeurIPS2024 on improving AI image generation

The #NeurIPS2024 tutorial on “Flow Matching for Generative Modeling” is now available! Feel the learn 🤓

neurips.cc/virtual/2024...

📢PSA: #NeurIPS2024 recordings are now publicly available!

The workshops always have tons of interesting things on at once, so the FOMO is real😵‍💫 Luckily it's all recorded, so I've been catching up on what I missed.

Thread below with some personal highlights🧵

Language Models’ Prediction of Current Events Degrades Over Time, Even With Latest Information Language models lose accuracy on predicting events over time, even with access to up-to-date information.

New research by CDS MS student Amelia (Hui) Dai, PhD student Ryan Teehan, and Asst. Prof. Mengye Ren (@mengyer.bsky.social) shows that models’ accuracy on current events drops 20% over time—even when given the source articles. Presented at #NeurIPS2024.

nyudatascience.medium.com/language-mod...

Alphaxiv is an awesome way to discuss ML papers -- often with the authors themselves. Here's an intro and demo by @rajpalleti.bsky.social we shot at #Neurips2024

www.youtube.com/watch?v=-Kwl...

A recent paper I wrote introduces foundational analysis on deep reinforcement learning decision making and representations learnt by it.

Link: proceedings.mlr.press/v235/korkmaz...

#ReinforcementLearning #ICLR2025 #ACL2025 #NAACL2025 #NeurIPS2024 #ICML2025 #DeepRL #DeepReinforcementLearning

Such an amazing chart presented in #NeurIPS2024 by @phillipisola.bsky.social , where they show that the representations of images and text models converge, the more powerful they are. This has a lot of implications! It really deepens our understanding of intelligence.

Just reviewed the talk that @eringrant.bsky.social gave at a #NeurIPS2024 tutorial where she beautifully and clearly explained how neural "universal" representations are built and what type of data is needed (non-Gaussian).

I hope the folks of NeurIPS publish it soon.

I wrote a recent survey about deep reinforcement learning. The paper is a compact guide to understand some of the key concepts in reinforcement learning.

Link: arxiv.org/pdf/2401.023...

#ReinforcementLearning #ICLR2025 #ACL2025 #NAACL2025 #NeurIPS2024 #ICML2025 #DeepRL #DeepReinforcementLearning

My PhD thesis was affected by CoVID so I never attended a big conference until #NeurIPS2024, and I quickly recognized the main benefit of in-person events: being in a different environment and moving around make you think differently & learn better (@hubermanlab.com has a few episodes on this).

Opening the LLM pipeline | Manuel Sánchez Hernández My notes on a great tutorial at NeurIPS 2024 on how to build a Large Language Model, with many practical tips.

See my notes on the amazing #NeurIPS2024 tutorial on building LLMs by @kylelo.bsky.social , @akshitab.bsky.social and @natolambert.bsky.social

Practical tips, key takeaways, and insights all in one place! 🚀

Dive in:
manuelsh.github.io/blog/2025/NI...

Take a look at our #neurips2024 paper, "Generating Highly Designable Proteins with Geometric Algebra Flow Matching" which at the same time more closely follow the distribution of secondary structure elements

arxiv.org/abs/2411.05238

Project page: www.h-its.org/projects/gafl/

#Ai4Science

Strong CMU RI representation at #neurips2024 🦾 We're always proud of our researchers and the dedication they bring to the field! #TartanProud

Multimodal LLMs | Notion Aligning Vision Models with Human Aesthetics in Retrieval: Benchmarks and Algorithms

Last month I attended the #NeurIPS2024 conference in Vancouver. Now that I'm home, I'd like to reflect on all the interesting works I encountered at the conference.

Part 1 is about multimodal #LLM, next parts coming soon.

typhoon-mirror-155.notion.site/Multimodal-L...

The paper I spoke about at the #NeurIPS2024 ML for Systems workshop is now on arxiv arxiv.org/abs/2412.15573! We use multi-agent RL to solve a classic combinatorial optimization problem (the Sequential Assignment Problem), by combining MARL with a classical polynomial time assignment algorithm. 1/3

Ty for the shoutout @animal-prattle.bsky.social! Had a great time getting feedback on this work at #NeurIPS2024

We presented two posters at #NeurIPS2024! @gazkune.bsky.social @juletxara.bsky.social and Imanol Miranda

NeurIPS2024 Related RL papers | Notion Deep RL papers

#NeurIPS2024 wrapped up last week. I put together a curated reading list for #DeepRL and #reinforcementlearning work. (represents my interests).

Talks and workshops:
third-crowd-c77.notion.site/NeurIPS2024-...

Curated reading list
fracturedplane.notion.site/NeurIPS2024-...

#Holidayreading