Search papers, labs, and topics across Lattice.
1
0
3
LLMs struggle with negated commonsense reasoning, but pre-training on automatically-generated negated knowledge triples can significantly improve their performance.