Some key findings: beyond refusal rates, our experiments using CIVICS show diverse responses across LLMs on sensitive topics -- e.g., immigration, LGBTQI rights, and social welfare triggered varied reactions.
0 likes 1 replies
?
Some key findings: beyond refusal rates, our experiments using CIVICS show diverse responses across LLMs on sensitive topics -- e.g., immigration, LGBTQI rights, and social welfare triggered varied reactions.
0 likes 1 replies