Search papers, labs, and topics across Lattice.
1
0
3
RLHF may create the illusion of neutrality in language models while leaving their partisan biases intact, posing risks for real-world applications.
