Post-Training Doesn't Create Your Model's Character. It Inherits One
By
Madhu Shantan
4mo ago
Source
MaximPost-Training Doesn't Create Your Model's Character. It Inherits Onemaxim-blog.ghost.ioIntroduction Every team building on top of LLMs has a version of the same mental model: pretraining teaches the model what it knows, and post-training teaches it how to behave. Don't want it to say harmful things? Train that out. Want it to be more helpful? Push
You might also wanna read
Reinforcement Pre-Training
arxiv.org·1y ago
Study Finds AI Discourse in Pretraining Data Creates Self-Fulfilling (Mis)alignment in LLMs
This research paper presents the first controlled study of how pretraining corpora containing discourse about AI systems causally influences
arxiv.org·1mo ago
Training Large Language Models for Honesty Through Self-Reported Confessions
Researchers propose a novel method to train large language models (LLMs) to be more honest by eliciting 'confessions' - self-reported accoun
arxiv.org·6mo agoBeyond the Easy-Button: Why Onboarding Our New AI Teammate is an Operational Evolution, Not an IT Project (via Passle)
digital-leadership-associates.passle.net·29d ago
Stabilizing LLM Behavior: The Assistant Axis Approach to Preventing Harmful Persona Drift
The article discusses how large language models (LLMs) develop character personas during training and introduces the concept of an "Assistan
anthropic.com·19d agoStabilizing LLM Behavior: The Assistant Axis Approach to Preventing Harmful Persona Drift
The article discusses how large language models (LLMs) develop character personas during training and introduces the concept of an "Assistan
anthropic.com·19d agoWhy Traditional Methods Are Better Than LLMs for Learning Programming
The article argues against using Large Language Models (LLMs) for learning programming, emphasizing that traditional learning methods and ha
agentultra.com·5mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.