Tracing Subliminal Learning Through Pretraining

·LessWrong··

This post summarizes our paper, Tracing Subliminal Learning Through Pretraining, which was developed as part of EleutherAI’s SOAR program. TL;DRWe study subliminal learning, the phenomenon in which a student model acquires a teacher’s preference through training data that appear semantically unrelated to that preference. Our work investigates subliminal learning through cross-model distillation, using the PolyPythia and DataDecide OLMo models to show that a teacher’s animal preference transfers ...

Read full article →

Related Articles

A 10-year-old CNN outperforms frontier LLMs for skin cancer diagnosis
sungam · Hacker News · 3h ago
Samsung running a TensorFlow lite model on 960x540 image to track you
mashally · Hacker News · 10h ago
Why is AI generated language so hard to understand?
canjobear · Hacker News · 1d ago
Every AI model has the same "controversial" take on humans
rakannimer · Hacker News · 3d ago
Five Qwen3.8-27B models match Fable 5 Using GVS5H Harness
rzk · Hacker News · 16h ago