Alignment faking in large language models
Paper
•
2412.14093
•
Published
•
9
Buck Shlegeris
bshlgrs
AI & ML interests
None yet
Organizations
authored
a
paper
about 1 year ago
authored
3
papers
almost 2 years ago
AI Control: Improving Safety Despite Intentional Subversion
Paper
•
2312.06942
•
Published
Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small
Paper
•
2211.00593
•
Published
•
2
Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
Paper
•
2401.05566
•
Published
•
30