A new paper titled “RLTF: Reinforcement Learning from Unit Test Feedback” proposes a novel method for…
Category: ARTICLE
Articles and other larger forms like tutorials and analysis for anyone wanting to learn more about how AI is progressing.
Meta AI Releases Code Llama Models for Advanced Code Generation
Meta AI has released Code Llama, a family of large language models for code that establishes…
Bringing Children’s Drawings to Life with AI
A team of researchers from Meta AI, Tencent, MIT, and Carnegie Mellon University have developed a…
Training Code Models to Follow Instructions with OctoPack
A new paper titled “OctoPack: Instruction Tuning Code Large Language Models” proposes a novel method for…
New AI System Accelerates Large Language Model Serving
A new artificial intelligence system called SpecInfer can significantly accelerate the speed of large language model…
GPT-4 is Easily Tricked with Encrypted Messages
A new study reveals a surprising vulnerability in large language models like GPT-4 – they can…
WizardMath: Empowering Large Language Models for Mathematical Reasoning
A team of researchers from Microsoft and the Shenzhen Institute of Advanced Technology have developed a…
Bayesian Flow Networks: A New Deep Generative Modeling Approach
A new deep generative modeling technique called Bayesian Flow Networks (BFNs) was recently introduced in a…
GPT-4 Code Interpreter Shows Impressive Math Reasoning Abilities With Self-Verification
A new study reveals that OpenAI’s latest version of GPT-4, known as the GPT-4 Code Interpreter,…
GPT-4 Outperforms RL Algorithms in Crafter by Reading Paper and Reasoning
A new approach called SPRING allows large language models (LLMs) like GPT-4 to achieve strong performance…
Humpack: Self-Alignment Allows Language Models to Improve Without Human Supervision
A new technique called “instruction backtranslation” allows large language models (LLMs) to improve their ability to…
New Benchmark Tests AI Agents in Real-World Challenges
Researchers from Tsinghua University, The Ohio State University, and UC Berkeley have introduced AgentBench, a new…
Scientists Invent AI That Can ‘Learn by Doing’ to Execute Computer Tasks
Researchers from UC Irvine and Carnegie Mellon University have developed a novel technique that allows AI…
Seeing Through the Brain: Reconstructing Visual Images from Brain Signals
A team of researchers from Shanghai Jiao Tong University and Microsoft Research have developed a new…
GPT-4 Still Unable to Reason, New Study Finds
A new study published on the preprint server Preprints.org argues that despite impressive advances, GPT-4 still…
Reinforcing Large Language Models with Retrospective Policy Optimization
Recent months have seen the rise of powerful new autonomous language agents built on top of…
The Quest to Overcome Key Challenges in Large Language Models
Large language models (LLMs) have rapidly risen to prominence, demonstrating impressive capabilities on a range of…
New Soft Mixture-of-Experts Model Sets New Benchmarks for Image Classification
A new paper from researchers at Google DeepMind proposes Soft Mixture-of-Experts (Soft MoE), a novel sparse…
Microsoft Unveils DeepSpeed-Chat to Democratize Training of Large Conversational AI Models
DeepSpeed-Chat is a new system introduced by Microsoft Researchers to make training large conversational AI models…
Uncovering How AI Masters New Senses
A new study from MIT CSAIL reveals how large language models like GPT-3 learn to integrate…