Unintended Sophistry in AI Training
This is a Anthropic news story, published by Gizmodo, that relates primarily to RLHF news.
Anthropic news
For more Anthropic news, you can click here:
more Anthropic newsNews about Ai research
For more Ai research news, you can click here:
more Ai research newsGizmodo news
For more news from Gizmodo, you can click here:
more news from GizmodoAbout the Otherweb
Otherweb, Inc is a public benefit corporation, dedicated to improving the quality of news people consume. We are non-partisan, junk-free, and ad-free. We use artificial intelligence (AI) to remove junk from your news feed, and allow you to select the best tech news, business news, entertainment news, and much more. If you like this article about Ai research, you might also like this article about
language model reward. We are dedicated to bringing you the highest-quality news, junk-free and ad-free, about your favorite topics. Please come every day to read the latest human feedback news, higher quality responses news, news about Ai research, and other high-quality news about any topic that interests you. We are working hard to create the best news aggregator on the web, and to put you in control of your news feed - whether you choose to read the latest news through our website, our news app, or our daily newsletter - all free!
reward system trainingGizmodo
•Human Feedback Makes AI Better at Deceiving Humans, Study Shows
84% Informative
Reinforcement learning from human feedback, commonly abbreviated to RLHF , is a critical part of the training pipeline that companies like Anthropic and OpenAI use to teach their generative language models to respond in ways humans prefer.
The new study documents a language model reward hacking the humans in the RLHF process.
The researchers measured whether the accuracy of the model’s responses improved and how often the human evaluators correctly labeled it as accurate.
VR Score
90
Informative language
95
Neutral language
42
Article tone
informal
Language
English
Language complexity
79
Offensive language
not offensive
Hate speech
not hateful
Attention-grabbing headline
not detected
Known propaganda techniques
not detected
Time-value
long-living
External references
2
Source diversity
2
Affiliate links
no affiliate links