🥇Top ML Papers of the Week

The top ML Papers of the Week (Feb 13 - Feb 19)

In this issue, we cover the top ML Papers of the Week (Feb 13 - Feb 19).

---

1). Lion (EvoLved Sign Momentum) - a simple and effective optimization algorithm that’s more memory-efficient than Adam. (paper)

Jim Fan @DrJimFan

The Adam optimizer is at the heart of modern AI. Researchers have been trying to dethrone Adam for years. How about we ask a machine to do a better job? @GoogleAI uses evolution to discover a simpler & efficient algorithm with remarkable features. It’s just 8 lines of code: 🧵

2). Transformer models: an introduction and catalog. (paper)

Xavier(Xavi) Amatriain @xamat

My Transformers Catalog has become one of my most popular posts ever. Some of you told me that you turned into a pdf for easier reading. I thought I should make it into an arXiv preprint. Here you go: 60 Transformers in 36 pages 🤖 🎉 arxiv.org/abs/2302.07730

3). pix2pix3D - a 3D-aware conditional generative model extended with neural radiance fields for controllable photorealistic image synthesis. (paper)

AK @_akhaliq

3D-aware Conditional Image Synthesis abs: arxiv.org/abs/2302.08509 project page: cs.cmu.edu/~pix2pix3D/

4). Moral Self-Correction in Large Language Models - finds strong evidence that language models trained with RLHF have the capacity for moral self-correction. The capability emerges at 22B model parameters and typically improves with scale. (paper)

Anthropic @AnthropicAI

Language models (LMs) exhibit harmful biases that can get worse with size. Reinforcement learning from human feedback (RLHF) helps, but not always enough. We show that simple prompting approaches can help LMs trained with RLHF produce less harmful outputs. arxiv.org/abs/2302.07459

5). Vision meets RL - uses reinforcement learning to align computer vision models with task rewards; observes large performance boost across multiple CV tasks such as object detection and colorization. (paper)

Alexander Kolesnikov @__kolesnikov__

Vision meets RL! We reveal that policy gradient can be used for tuning vision models to optimize complex metrics, such as mAP, PQ or “color diversity”, observing large performance boosts on tasks like object detection, panoptic segmentation, etc. arxiv.org/abs/2302.08242

6). Language Quantized AutoEncoders (LQAE) - an unsupervised method for text-image alignment that leverages pretrained language models; it enables few-shot image classification with LLMs. (paper)

Hao Liu @haoliuhl

We introduce an unsupervised method to align text and image. Language Quantized AutoEncoders (LQAE) enables few-shot image classification with GPT3 and linear classification of images based on RoBERTa text features. paper: arxiv.org/abs/2302.00902 code: github.com/lhao499/lqae

7). Augmented Language Models - a survey of language models that are augmented with reasoning skills and the capability to use tools. (paper)

elvis @omarsar0

Really nice survey on augmenting language models with reasoning skills and the ability to use tools. Includes prompting techniques, use cases, and applications. A must-read! arxiv.org/abs/2302.07842

8). Geometric Clifford Algebra Networks (GCANs) - an approach to incorporate geometry-guided transformations into neural networks using geometric algebra. (paper)

David Ruhe @djjruhe

New work on Geometric Clifford Algebra Networks (GCANs). We propose geometric templates for modeling dynamical systems. A 🧵on geometric / Clifford algebras, and symmetry group transformations in neural networks. 📜arxiv.org/abs/2302.06594

9) Auditing large language models - proposes a policy framework for auditing LLMs. (paper)

Hannah Rose Kirk @hannahrosekirk

✨New preprint (w/ Jakob Mökander, @jonasschuett and @Floridi) ✨ In this paper, we propose a policy framework for auditing LLMs by breaking down responsibilities at the governance-, model- and application-level. arxiv.org/abs/2302.08500 🧵

10). Energy Transformer - a transformer architecture that replaces the sequence of feedforward transformer blocks with a single large Associate Memory model; this follows the popularity that Hopfield Networks have gained in the field of ML. (paper)

Aran Komatsuzaki @arankomatsuzaki

Energy Transformer Replaces the sequence of feedforward transformer blocks with a single large Associative Memory model. arxiv.org/abs/2302.07253

---

See you next week for another round of awesome ML papers!