Where the Goblins Came From

openai.com · · Covered in 13 articles from 12 sources

Simulating Simulators

lesswrong.com·

Predictive Data Debugging: Reveal and Shape What Your Model Learns, Before You Train (11 minute read)

Appreciating Exif

brentfitzgerald.com·

Agents Need Work Data: A Primer on RLWD, or Reinforcement Learning on Work Data

anjalishriva.com··Hacker News

How to Train Your Goblin

goblins.mchen.workers.dev··Hacker News, Hacker News

How Adversaries Use AI, Skill Issues, Using IDEs for C2

Claude, Author of the Humanitas

greaterwrong.com··Hacker News

Claude, Author of the Humanitas

linch.substack.com··Substack

Claude, Author of the Humanitas

lesswrong.com··Hacker News

Generalization Dynamics of LM Pre-Training

jiaxin-wen.github.io··Hacker News

Fixing LLM Writing with Distribution Fine Tuning

rosmine.ai··Hacker News, Hacker News

Links For You (5/17/26)

raymondcamden.com·

In other languages

«Боргворглер: боргает ворглер», или как сделать действительно полезную документацию к проекту