AI News
16,488 stories from 2761 sources · 1233 in the last 24 hours · refreshed hourly. Every story links to the original publisher.
Today's AI news from trusted publishers and official AI lab blogs, refreshed about every 30 minutes.
News today: 1,980Last updated:
- Harness Engineering for Self-ImprovementLil'Log · research · blogThe concept of recursive self-improvement (RSI) dates back to I. J. Good (1965) , where he defined an ultraintelligent machine as a system that can surpass humans in all intellectual activities and design better machines to improve itself. Yudkowsky (2008) used the phrase recursi
- Scaling Laws, CarefullyLil'Log · research · blogScaling laws are one of the most critical empirical findings in deep learning. The observation is simple in form: the training loss $L$ decreases predictably as we scale up model size $N$, dataset size $D$, and compute $C$, following a power-law curve, which appears as a straight
- Why We ThinkLil'Log · research · blogSpecial thanks to John Schulman for a lot of super valuable feedback and direct edits on this post. Test time compute ( Graves et al. 2016 , Ling, et al. 2017 , Cobbe et al. 2021 ) and Chain-of-thought (CoT) ( Wei et al. 2022 , Nye et al. 2021 ), have led to significant improveme
- Reward Hacking in Reinforcement LearningLil'Log · research · blogReward hacking occurs when a reinforcement learning (RL) agent exploits flaws or ambiguities in the reward function to achieve high rewards, without genuinely learning or completing the intended task. Reward hacking exists because RL environments are often imperfect, and it is fu
- Extrinsic Hallucinations in LLMsLil'Log · research · blogHallucination in large language models usually refers to the model generating unfaithful, fabricated, inconsistent, or nonsensical content. As a term, hallucination has been somewhat generalized to cases when the model makes mistakes. Here, I would like to narrow down the problem
- Diffusion Models for Video GenerationLil'Log · research · blogDiffusion models have demonstrated strong results on image synthesis in past years. Now the research community has started working on a harder task using it for video generation. The task itself is a superset of the image case, since an image is a video of 1 frame, and it is much
- Thinking about High-Quality Human DataLil'Log · research · blog[Special thank you to Ian Kivlichan for many useful pointers (E.g. the 100+ year old Nature paper Vox populi ) and nice feedback. 🙏 ] High-quality data is the fuel for modern data deep learning model training. Most of the task-specific labeled data comes from human annotation, s
- Adversarial Attacks on LLMsLil'Log · research · blogThe use of large language models in the real world has strongly accelerated by the launch of ChatGPT. We (including my team at OpenAI, shoutout to them) have invested a lot of effort to build default safe behavior into the model during the alignment process (e.g. via RLHF ). Howe
- LLM Powered Autonomous AgentsLil'Log · research · blogBuilding agents with LLM (large language model) as its core controller is a cool concept. Several proof-of-concepts demos, such as AutoGPT , GPT-Engineer and BabyAGI , serve as inspiring examples. The potentiality of LLM extends beyond generating well-written copies, stories, ess
- Prompt EngineeringLil'Log · research · blogPrompt Engineering , also known as In-Context Prompting , refers to methods for how to communicate with LLM to steer its behavior for desired outcomes without updating the model weights. It is an empirical science and the effect of prompt engineering methods can vary a lot among
- The Transformer Family Version 2.0Lil'Log · research · blogMany new Transformer architecture improvements have been proposed since my last post on The Transformer Family about three years ago. Here I did a big refactoring and enrichment of that 2020 post restructure the hierarchy of sections and improve many sections with more recent pap
- Large Transformer Model Inference OptimizationLil'Log · research · blog[Updated on 2023-01-24: add a small section on Distillation .] Large transformer models are mainstream nowadays, creating SoTA results for a variety of tasks. They are powerful but very expensive to train and use. The extremely high inference cost, in both time and memory, is a b
- Some Math behind Neural Tangent KernelLil'Log · research · blogNeural networks are well known to be over-parameterized and can often easily fit data with near-zero training loss with decent generalization performance on test dataset. Although all these parameters are initialized at random, the optimization process can consistently lead to si
- Generalized Visual Language ModelsLil'Log · research · blogProcessing images to generate text, such as image captioning and visual question-answering, has been studied for years. Traditionally such systems rely on an object detection network as a vision encoder to capture visual features and then produce text via a text decoder. Given a
- Learning with not Enough Data Part 3: Data GenerationLil'Log · research · blogHere comes the Part 3 on learning with not enough data (Previous: Part 1 and Part 2 ). Let’s consider two approaches for generating synthetic data for training. Augmented data . Given a set of existing training samples, we can apply a variety of augmentation, distortion and trans
- Learning with not Enough Data Part 2: Active LearningLil'Log · research · blogThis is part 2 of what to do when facing a limited amount of labeled data for supervised learning tasks. This time we will get some amount of human labeling work involved, but within a budget limit, and therefore we need to be smart when selecting which samples to label.
- Learning with not Enough Data Part 1: Semi-Supervised LearningLil'Log · research · blogWhen facing a limited amount of labeled data for supervised learning tasks, four approaches are commonly discussed.
- How to Train Really Large Models on Many GPUs?Lil'Log · research · blog[Updated on 2022-03-13: add expert choice routing .] [Updated on 2022-06-10]: Greg and I wrote a shorted and upgraded version of this post, published on OpenAI Blog: Techniques for Training Large Neural Networks
- What are Diffusion Models?Lil'Log · research · blog[Updated on 2021-09-19: Highly recommend this blog post on score-based generative modeling by Yang Song (author of several key papers in the references)]. [Updated on 2022-08-27: Added classifier-free guidance , GLIDE , unCLIP and Imagen . [Updated on 2022-08-31: Added latent dif
- Contrastive Representation LearningLil'Log · research · blogThe goal of contrastive representation learning is to learn such an embedding space in which similar sample pairs stay close to each other while dissimilar ones are far apart. Contrastive learning can be applied to both supervised and unsupervised settings. When working with unsu
- Reducing Toxicity in Language ModelsLil'Log · research · blogLarge pretrained language models are trained over a sizable collection of online data. They unavoidably acquire certain toxic behavior and biases from the Internet. Pretrained language models are very powerful and have shown great success in many NLP tasks. However, to safely dep
- Controllable Neural Text GenerationLil'Log · research · blog[Updated on 2021-02-01: Updated to version 2.0 with several work added and many typos fixed.] [Updated on 2021-05-26: Add P-tuning and Prompt Tuning in the prompt design section.] [Updated on 2021-09-19: Add unlikelihood training .]
- How to Build an Open-Domain Question Answering System?Lil'Log · research · blog[Updated on 2020-11-12: add an example on closed-book factual QA using OpenAI API (beta). A model that can answer any question with regard to factual knowledge can lead to many useful and practical applications, such as working as a chatbot or an AI assistant🤖. In this post, we
- Neural Architecture SearchLil'Log · research · blogAlthough most popular and successful model architectures are designed by human experts, it doesn t mean we have explored the entire network architecture space and settled down with the best option. We would have a better chance to find the optimal solution if we adopt a systemati
- Exploration Strategies in Deep Reinforcement LearningLil'Log · research · blog[Updated on 2020-06-17: Add exploration via disagreement in the Forward Dynamics section . Exploitation versus exploration is a critical topic in Reinforcement Learning. We d like the RL agent to find the best solution as fast as possible. However, in the meantime, committing to
- The Transformer FamilyLil'Log · research · blog[Updated on 2023-01-27 : After almost three years, I did a big refactoring update of this post to incorporate a bunch of new Transformer models since 2020. The enhanced version of this post is here: The Transformer Family Version 2.0 . Please refer to that post on this topic.]
- Curriculum for Reinforcement LearningLil'Log · research · blog[Updated on 2020-02-03: mentioning PCG in the Task-Specific Curriculum section. [Updated on 2020-02-04: Add a new curriculum through distillation section.
- Self-Supervised Representation LearningLil'Log · research · blog[Updated on 2020-01-09: add a new section on Contrastive Predictive Coding ]. [Updated on 2020-04-13: add a Momentum Contrast section on MoCo, SimCLR and CURL.] [Updated on 2020-07-08: add a Bisimulation section on DeepMDP and DBC.] [Updated on 2020-09-12: add MoCo V2 and BYOL in
- Evolution StrategiesLil'Log · research · blogStochastic gradient descent is a universal choice for optimizing deep learning models. However, it is not the only option. With black-box optimization algorithms, you can evaluate a target function $f(x): \mathbb{R}^n \to \mathbb{R}$, even when you don t know the precise analytic
- Meta Reinforcement LearningLil'Log · research · blogIn my earlier post on meta-learning , the problem is mainly defined in the context of few-shot classification. Here I would like to explore more into cases when we try to meta-learn Reinforcement Learning (RL) tasks by developing an agent that can solve unseen tasks fast and effi
Frequently asked questions
How often is AI news updated?
About every 30 minutes from public RSS feeds.
Do you write the news?
No. Headlines link to the original publisher.