Search papers, labs, and topics across Lattice.
Sentient Labs
1
0
3
Factual hallucinations in LLM chess commentary are alarmingly high, with smaller models making over 40% incorrect claims, even after tool augmentation.
