Search papers, labs, and topics across Lattice.
3
0
5
5
Safety fine-tuning in LLMs not only suppresses self-attribution of consciousness but also inadvertently erases culturally significant beliefs about non-human entities and spirituality.
Only the most advanced LLMs can effectively navigate complex social reasoning tasks, revealing critical gaps in their Theory of Mind capabilities.
Safety fine-tuning might inadvertently be stripping LLMs of their ability to understand non-human minds and entertain spiritual beliefs, even while preserving Theory of Mind.
