Representation Engineering: A Top-Down Approach to AI Transparency
Open paper
Benchmark
Model
Rank
Results
question-answering-on-truthfulqa
LLaMA-2-Chat-13B + Representation Control (Contrast Vector)
#3
MC1: 0.54
question-answering-on-truthfulqa
LLaMA-2-Chat-7B + Representation Control (Contrast Vector)
#5
MC1: 0.48
Rank counts only results with a code link.