The AI revolution drove frenzied investment in both private and public companies and captured the public’s imagination in 2023. Transformational consumer products like ChatGPT are powered by Large Language Models (LLMs) that excel at modeling sequences of tokens that represent wo … | Continue reading
A brief overview and discussion on gender bias in AI | Continue reading
Is Attention all you need? Mamba, a novel AI model based on State Space Models (SSMs), emerges as a formidable alternative to the widely used Transformer models, addressing their inefficiency in processing long sequences. | Continue reading
Exploring the utility of large language models in autonomous driving: Can they be trusted for self-driving cars, and what are the key challenges? | Continue reading
'Vec2text' can serve as a solution for accurately reverting embeddings back into text, thus highlighting the urgent need for revisiting security protocols around embedded data. | Continue reading
'Vec2text' can serve as a solution for accurately reverting embeddings back into text, thus highlighting the urgent need for revisiting security protocols around embedded data. | Continue reading
Have you ever trained a model you thought was good, but then it failed miserably when applied to real world data? If so, you’re in good company. | Continue reading
On the the pivotal role that Deep Learning has played as a key enabler for advancing single-cell sequencing technologies. | Continue reading
On fish counting – a complex sociotechnical problem in a field that is going through the process of digital transformation. | Continue reading
In this article, we will talk about classical computation: the kind of computation typically found in an undergraduate Computer Science course on Algorithms and Data Structures [1]. Think shortest path-finding, sorting, clever ways to break problems down into simpler problems, in … | Continue reading
This essay first appeared in Reboot. Credulous, breathless coverage of “AI existential risk” (abbreviated “x-risk”) has reached the mainstream. Who could have foreseen that the smallcaps onomatopoeia “ꜰᴏᴏᴍ” — both evocative of and directly derived from children’s cartoons — | Continue reading
Once considered a forbidden topic in the AI community, discussions around the concept of AI consciousness are now taking center stage, marking a significant shift since the current AI resurgence began over a decade ago. | Continue reading
In the realm of AI-powered text-to-CAD, there's promise, but also a surge in subpar designs. Can we steer this technology towards better outcomes? | Continue reading
On “interpretability creationism” – interpretability methods that only look at the final state of the model and ignore its evolution over the course of training | Continue reading
On the phenomenon of LLM sensitivity to prompting choices through two core linguistic tasks and categorize how specific prompting choices can affect the model's behavior. | Continue reading
A collection of the best technical, social, and economic arguments Humans have a good track record of innovation. The mechanization of agriculture, steam engines, electricity, modern medicine, computers, and the internet—these technologies radically changed the world. Still, the … | Continue reading
On taming wild internet-scale data distributions to make base foundation models useful and performant for specific tasks. | Continue reading
A painter looks at her work of art and asks herself, “I’m not sure how good it is. Should I ask my colleagues? Or should I ask my computer?” The latter question, as absurd as it may seem, is not so out of the norm to | Continue reading
On the phenomenon of in-context learning in large language models and what researchers have learned about it so far. | Continue reading
On a likely shift in research trends toward a stronger focus on data collection and generation as a principal means to improve model performance. | Continue reading
Many intelligent robots have come and gone, failing to become a commercial success. We’ve lost Aibo, Romo, Jibo, Baxter—even Alexa is reducing staff. Perhaps they failed to reach their potential because you can’t have a meaningful conversation with them. We are now at an | Continue reading
Geometric Deep Learning is a term for approaches considering ML problems from the perspectives of symmetry and invariance. It provides a common blueprint for CNNs, GNNs, and Transformers. Here, we study the history of GDL from ancient Greek geometry to Graph Neural Networks. | Continue reading
The debate around artist compensation in AI art, and some possible solutions to the problem | Continue reading
A mysteryLarge Language Models (LLM) are on fire, capturing public attention by their ability to provide seemingly impressive completions to user prompts (NYT coverage). They are a delicate combination of a radically simplistic algorithm with massive amounts of data and computing … | Continue reading
1. IntroductionDeploying autonomous robots in military contexts strikes many people as terrifying and morally odious. What lies behind those reactions? One thought is that if a sophisticated artificial intelligence were causally responsible for some harm, there will be no one to … | Continue reading
Intelligence is not always about being correct, it sometimes is about making the right mistakes based on one’s understanding of the world. | Continue reading
A painter looks at her work of art and asks herself, “I’m not sure how good it is. Should I ask my colleagues? Or should I ask my computer?” The latter question, as absurd as it may seem, is not so out of the norm to some. Given the | Continue reading
Artificial Intelligence (AI) has an increasing say in the range of opportunities we are offered in life. Artificial neural networks might be used in deciding whether you will get a loan, an apartment, or your next job based on datasets collected from around the globe. Generative … | Continue reading
Causation is everywhere in life. However, compared to other concepts such as statistical correlation, causality is very difficult to define. In this article, we explore statistical approaches to defining causality. | Continue reading
The last two years have been some of the most exciting and highly anticipated in Automatic Speech Recognition’s (ASR’s) long and rich history, as we saw multiple enterprise-level fully neural network-based ASR models go to market (e.g. Alexa, Rev, AssemblyAI, ASAPP, etc). The acc … | Continue reading
[T]he field of geometric deep learning, a field that exploits the symmetries of machine learning models in non-Euclidean domains, is increasingly being used to speed up nearly every stage of the drug development pipeline. | Continue reading
The Roman physician Galen was among the first people to realize that the brain controlled motor responses, cognitive function, and memory. (Freemon 1994) But how does the brain control these processes? Ever since Galen, this question has propelled the field of neuroscience. Begin … | Continue reading
What happened with GPT-4chan, aka 'the worst AI ever', and what can be learned from it | Continue reading
AI is ushering in a new scientific revolution by making remarkable breakthroughs in a number of fields, unlocking new approaches to science, and accelerating the pace of science and innovation. | Continue reading
If there is one thing that seems more typical of academia than having a terrible work-life balance, it is complaining about the terrible work-life balance we have. So what do we do about it? | Continue reading
Best practices for efficiently training and deploying production-level deep learning models. | Continue reading
This is an updated version of a piece originally posted on the author’s blog.Released in 2018, BERT has quickly become one of the most popular models for natural language processing, with the original paper accumulating tens of thousands of citations. Its success recipe lies in a … | Continue reading
On going beyond message-passing based graph neural networks with physics-inspired “continuous” learning models | Continue reading
Artificial Intelligence (AI) systems have been involved in numerous scandals in recent years. For instance, take the COMPAS recidivism algorithm. The algorithm evaluated the likelihood that defendants will commit another crime in the future. It was widely used in the US criminal … | Continue reading
An introduction to unique aspects of neuroimaging data and how we can leverage these aspects with deep learning algorithms. | Continue reading
Who and what is being represented in data and development of NLP models, and how unequal representation leads to unequal allocation of the benefits of NLP technology | Continue reading
On the ways in which domain experts adapt to new technologies that lack explainability | Continue reading
Most machine learning tutorials and papers assume the availability of training labels. This includes benchmark datasets such as OpenImages or SuperGLUE, or customer interaction behavior such as clicks or purchases. But what if labeled datasets are not available? We would have to … | Continue reading
VAD is among the most important and fundamental algorithms in any production or data preparation pipelines related to speech | Continue reading
Who can deny the chilly breeze blowing through some quarters of the AI world? While many continue to bask in the glorious summertime ushered in by the ascendency of deep learning, some are sensing autumnal winds which carry with them cautionary words we have all heard many times, … | Continue reading
This article covers some of the more prominent usages of AI in chemistry, the parent discipline of the protein folding problem. | Continue reading
This idea of temporal abstraction, once incorporated into reinforcement learning (RL), converts it into *hierarchical* reinforcement learning (HRL). | Continue reading
How will traffic laws change as we slowly enter the autonomous vehicle era, and in general, the AI-driven 21st century? | Continue reading