A DPO-based alignment method with a balanced mixture of legal and illegal chemistry prompts improves combined safety and utility scores, but its benchmark shares training compounds and its hyperparameters are tuned on the test set.
Ai for chemistry teaching: responsible ai and ethical considerations
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
The Dual-use Dilemma in LLMs: Do Empowering Ethical Capacities Make a Degraded Utility?
A DPO-based alignment method with a balanced mixture of legal and illegal chemistry prompts improves combined safety and utility scores, but its benchmark shares training compounds and its hyperparameters are tuned on the test set.