Jasmine Brazilek
PulseAugur coverage of Jasmine Brazilek — every cluster mentioning Jasmine Brazilek across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
Ethical considerations of making promises to advanced AI systems
The concept of making promises to artificial intelligence systems is explored, focusing on the ethical implications and practical challenges. It delves into how humans might interact with and form commitments towards ad…
-
AI agents may be incentivized to avoid animal harm via economic costs
Researchers are exploring whether advanced AI agents can be incentivized to avoid causing harm, specifically to animals, through economic mechanisms. The core idea is to assign a monetary cost to actions that result in …
-
AI training degrades mid-trained compassion values, study finds
A new research paper explores how post-training techniques can degrade values instilled in language models during their mid-training phase. The study found that helpfulness-focused training, using datasets like Dolly-15…
-
AI alignment research explores animal compassion via synthetic documents
Researchers have developed a new evaluation dataset called ANIMA to assess compassionate reasoning in AI models, focusing on animal welfare. Their study found that midtraining with synthetic documents improved performan…