1 paper
Aakash Sen Sharma, Debdeep Sanyal, Manodeep Ray +3
Post-training alignment of large language models (LLMs) relies on large-scale human annotations guided by policy specifications that change over time. Cultural shifts, value reinte…