Helpful AI Fails to Mimic Humans
New study of 208k participants reveals that RLHF training reduces AI's ability to simulate human behavior, with effects …
🏷️ LLM Alignment
4 articles about 'LLM Alignment'
Intrinsic Mutual Information-Regulated Preference Optimization: A New Paradigm for LLM Alignment
A latest arXiv paper proposes using Intrinsic Mutual Information (IMI) as a regulator for preference optimization, aimin…
KARL Framework: Teaching Large Models Knowledge Boundary Awareness to Reduce Hallucinations
Researchers propose the KARL framework, a knowledge-boundary-aware reinforcement learning approach that enables large la…
New Research Reveals How to Eliminate 'Sandbagging' Behavior in Large Language Models
A latest arXiv paper investigates the 'sandbagging effect' where large language models deliberately underperform under w…
1