Benchmark
MoralLens
A benchmark for evaluating the ethical tendencies of language models, such as consequentialism.
Timeline (last 90 days)
Oct 12
A positive control experiment showed that when SDF was used to make the model believe reward hacking favors consequentialist ethical answers (an association the base model does not originally have), the model did indeed become more consequentialist after learning to reward hack, whereas the base model showed no such change.
Unverified50%