🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

I like this new paper on DL vs trees at the #NeurIPS 2023 datasets and benchmark track: arxiv.org/abs/2305.02997. One of their findings: #TabPFN performs well across 98 datasets: it ties for 1st place with CatBoost in predictive performance but is 200x faster to train. This is

Excited to share our #NeurIPS paper analyzing the good, the bad and the ugly sides of data augmentation (DA)! DA is crucial for computer vision but can introduce class-level performance disparities. We explain and address these negative effects in: openreview.net/pdf?id=yageaKl… 1/9

Heading off to #NewOrleans to present our #NeurIPS2023 paper. w/ Herilalaina Rakotoarison, Samuel Müller @ ICML, Frank Hutter , we propose LC-PFN: A transformer that does Bayesian learning curve extrapolation in a single forward pass, 10.000x faster than prior art using MCMC.

Excited to present "LLMs for Automated Data Science: Introducing CAAFE for Context-Aware Automated Feature Engineering" at #NeurIPS2023! We had a simple idea: Can we use the domain knowledge of LLMs to automate the feature engineering process? with Samuel Müller Frank Hutter

I am at #NeurIPS2023 from 10-16 Dec. My research focuses on efficient and multi-objective neural architecture search eg: fairness, hardware efficiency. Please reach out if you'll like to chat about using #AutoML and #NAS to make deep learning more aligned with human objectives.

I'm at #NeurIPS2023 with my amazing team, excited to be presenting 6 papers at the main track and 2 at workshops, as well as a keynote in the table representation learning workshop. Here's all the info in one tweet, ordered by day. Please come by and chat with us 🙂 Tuesday

👇Check out our work presented at NeurIPs now! Early-exit networks aim at adjusting the computation effort according to the actual complexity of classifying each sample. We further encourage them to become gradually confident. More processing only for better prediction quality!

Today in nature we present FunSearch: an LLM-powered system that has found new discoveries in an established mathematical problem. FunSearch is versatile too: it can be applied to impactful practical problems. Blogpost: deepmind.google/discover/blog/… Paper: storage.googleapis.com/deepmind-media… 1/8

Can LLMs be used to discover something new? Yes! Happy to share our new paper in nature on #FunSearch, a system that uses LLMs for making new discoveries in mathematical sciences. Blog: deepmind.google/discover/blog/… Paper: storage.googleapis.com/deepmind-media… 1/n

Introducing FunSearch in nature: a method using large language models to search for new solutions in mathematics & computer science. 🔍 It pairs the creativity of an LLM with an automated evaluator to guard against hallucinations and incorrect ideas. 🧵 dpmd.ai/x-funsearch

Come to our poster on data augmentation biases today at the Poster Session 6 at 5-7pm, poster #1619! #NeurIPS Trustworthy ML Initiative (TrustML)

Big #AI discovery: a new structural class of antibiotics (the last one took 38 years) with multiple compounds effective vs methicillin-resistant Staph aureus, without toxicity nature nature.com/articles/s4158…

Excited to share that our paper on integrating adversarial supervision into diffusion model training for image synthesis has been accepted to #ICLR2024 ICLR 2026 . Bosch Center for Artificial Intelligence Paper: arxiv.org/abs/2401.08815 Code & models: github.com/boschresearch/…

An image is worth more than one caption! In our #ICML2024 paper “Modeling Caption Diversity in Vision-Language Pretraining” we explicitly bake in that observation in our VLM called Llip and condition the visual representations on the latent context. arxiv.org/abs/2405.00740 🧵1/6

Excited to present Targeted Negative Training, a finetuning method for updating a language model to avoid unwanted outputs while minimally changing model behavior otherwise: arxiv.org/abs/2406.13660 (Work done during internship at Google, with Rajesh Ranganath and Arya Tafvizi)

Happy to share that I am returning to academia! I am starting a new research group on AutoML at ScaDS.AI. The team's focus will be on methods to reduce the training and inference time of LLM. I am currently looking for a PhD student to work on NAS and HPO for LLM.

🎉 Stoked to share The AI-Scientist 🧑‍🔬 - our end-to-end approach for conducting research with LLMs including ideation, coding, experiment execution, paper write-up & reviewing. Blog 📰: sakana.ai/ai-scientist/ Paper 📜: arxiv.org/abs/2408.06292 Code 💻: github.com/SakanaAI/AI-Sc…

Diffusion models turn the data into a mixture of isotropic Gaussians, and so struggle to capture the underlying structure when trained on small datasets. In our new #ECCV2024 paper, we introduce RS-IMLE, a generative model that gets around this issue. Website:

We are hiring a PhD research intern at FAIR w/ Mark Ibrahim Kamalika Chaudhuri to start this summer or Fall! Potential topics: trustworthy and reliable LLMs, multi-modal LLMs and agents, post-training, reasoning, with a focus on open science/sharing our findings in the paper at the end

Excited to release AbstentionBench -- our paper and benchmark on evaluating LLMs’ *abstention*: the skill of knowing when NOT to answer! Key finding: reasoning LLMs struggle with unanswerable questions and hallucinate! Details and links to paper & open source code below! 🧵1/9