PT-ALIGN uses self-generated positive and toxic sample pairs with MLE plus fine-grained token-level unlikelihood training to improve LLM safety with minimal human annotation.
Self-alignment of large language models via monopolylogue-based social scene simulation,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Refining Positive and Toxic Samples for Dual Safety Self-Alignment of LLMs with Minimal Human Interventions
PT-ALIGN uses self-generated positive and toxic sample pairs with MLE plus fine-grained token-level unlikelihood training to improve LLM safety with minimal human annotation.