Record
- Authors
- Greenblatt, R.; Shlegeris, B.; Sachan, K.; Roger, F.
- Year
- 2023
- Title
- AI Control: Improving Safety Despite Intentional Subversion
- Published in
- arXiv
- Type
- Preprint
- DOI
- 10.48550/arxiv.2312.06942
- arXiv
- arXiv:2312.06942
- Links
- https://arxiv.org/abs/2312.06942
- Our note
- none yet
- Verification
- verified: its identifier resolves to a record with the same title and author (checked against OpenAlex, 2026-11-14).
Identifiers found: OpenAlex W4389714039 · DOI 10.48550/arxiv.2312.06942 · arXiv:2312.06942 - From the lookup
- doi (the pages' own wording is under "As cited" below)
- Id
greenblatt-2023-ai-control-improving-safety-despite-intentional
The pages disagree on year: year 2023 or 2024. The record above takes the most common value.
Cited on
- Automated Researchers Can Subtly Sandbag read status: not stated
- Sabotage Evaluations for Frontier Models read status: not stated
- Three Sketches of ASL-4 Safety Case Components read status: not read
As cited
The pages cite this work in 3 forms.
-
Greenblatt, R., Shlegeris, B., Sachan, K., and Roger, F. (2023). AI Control: Improving Safety Despite Intentional Subversion. arXiv:2312.06942. Cited above only as the paper describes it.
-
Greenblatt, R., Shlegeris, B., Sachan, K., and Roger, F. (2023). AI Control: Improving Safety Despite Intentional Subversion. arXiv:2312.06942. https://arxiv.org/abs/2312.06942
-
Greenblatt, R., Shlegeris, B., Sachan, K., and Roger, F. (2024). AI Control: Improving Safety Despite Intentional Subversion. ICML 2024. arXiv:2312.06942 (cited as the post cites it; not read for this note).
BibLaTeX
@misc{greenblatt-2023-ai-control-improving-safety-despite-intentional,
author = {Greenblatt, R. and Shlegeris, B. and Sachan, K. and Roger, F.},
title = {{AI Control: Improving Safety Despite Intentional Subversion}},
date = {2023},
doi = {10.48550/arxiv.2312.06942},
eprint = {2312.06942},
eprinttype = {arxiv}
}