Everything you care about in one place
Follow feeds: blogs, news, RSS and more. An effortless way to read and digest content of your choice.
Get Feederblog.ml.cmu.edu
Blog | Machine Learning | Carnegie Mellon University
Get the latest updates from Blog | Machine Learning | Carnegie Mellon University directly as they happen.
Follow now 130 followers
Latest posts
Last updated 23 days ago
Diffusion Beats Autoregressive in Data-Constrained Settings
23 days ago
TLDR If you are compute-constrained, use autoregressive models; if you are data-constrained,...
Verlog: A Multi-turn RL framework for LLM agents
about 1 month ago
Verlog is a multi-turn reinforcement learning framework built for long-horizon LLM-agentic tasks...
Carnegie Mellon University at ICML 2025
3 months ago
CMU researchers are presenting 127 papers at the Forty-Second International Conference on...
RLHF 101: A Technical Tutorial on Reinforcement Learning from Human Feedback
5 months ago
Reinforcement Learning from Human Feedback (RLHF) is a popular technique used to...
Unlearning or Obfuscating? Jogging the Memory of Unlearned LLMs via Benign Relearning
5 months ago
Machine unlearning is a promising approach to mitigate undesirable memorization of training...
Carnegie Mellon University at ICLR 2025
6 months ago
CMU researchers are presenting 143 papers at the Thirteenth International Conference on...
Allie: A Human-Aligned Chess Bot
6 months ago
Play against Allie on lichess! Introduction In 1948, Alan Turning designed what...
LLM Unlearning Benchmarks are Weak Measures of Progress
6 months ago
TL;DR: “Machine unlearning” aims to remove data from models without retraining the...
Copilot Arena: A Platform for Code
6 months ago
Figure 1. Copilot Arena is a VSCode extension that collects human preferences...
Optimizing LLM Test-Time Compute Involves Solving a Meta-RL Problem
9 months ago
Figure 1: Training models to optimize test-time compute and learn “how to...
Inductive biases of neural network modularity in spatial navigation
10 months ago
TL;DR: The brain may have evolved a modular architecture for daily tasks,...
Human-AI Collaboration in Physical Tasks
10 months ago
TL;DR: At SmashLab, we’re creating an intelligent assistant that uses the sensors...