Rais Grant
Proposer of a research team at Rome AI Safety (RAIS), funded by Coefficient Giving with $80k to develop disentangled, composable steering vectors for fine-grained behavioral control of LLMs.
Proposer of a research team at Rome AI Safety (RAIS), funded by Coefficient Giving with $80k to develop disentangled, composable steering vectors for fine-grained behavioral control of LLMs.