Show HN: I trained an AI to write my YouTube titles
🤖AI
Flag this post
🧠🚀 Excited to introduce Supervised Reinforcement Learning—a framework that leverages expert trajectories to teach small LMs how to reason through hard problems ...
threadreaderapp.com·3d
🤖AI
Flag this post
Model Inversion with Layer-Specific Modeling and Alignment for Data-Free Continual Learning
arxiv.org·2d
🧠Machine Learning
Flag this post
Automated Crack Mitigation in Laser Weld Repairs via Adaptive Thermal Gradient Optimization
🧠Machine Learning
Flag this post
How Data Mixing Shapes In-Context Learning: Asymptotic Equivalence for Transformers with MLPs
arxiv.org·3d
🤖Transformers
Flag this post
RePro: Training Language Models to Faithfully Recycle the Web for Pretraining
📝Natural Language Processing
Flag this post
HUME: Measuring the Human-Model Performance Gap in Text Embedding Task
📝Natural Language Processing
Flag this post
Supervised Reinforcement Learning: From Expert Trajectories to Step-wise Reasoning
arxiv.org·2d
🚀MLOps
Flag this post
Convexity Switching: The Secret to Faster, Smarter Neural Net Training?
👁️Computer Vision
Flag this post
[Open Source] We deployed numerous agents in production and ended up building our own GenAI framework
⛓️LangChain
Flag this post
Loading...Loading more...