Eliciting Latent Knowledge paper
GPTKB entity
Statements (24)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:academic_journal
|
| gptkbp:allows |
The paper discusses methods for eliciting knowledge that is latent in AI systems, with a focus on alignment and interpretability.
|
| gptkbp:arXivID |
2206.01753
|
| gptkbp:author |
gptkb:Paul_Christiano
gptkb:Jan_Leike gptkb:John_Schulman gptkb:Mark_Xu gptkb:Dario_Amodei gptkb:Sam_Bowman gptkb:Tom_Henighan gptkb:Jeff_Wu gptkb:Owain_Evans gptkb:Beth_Barnes gptkb:Evan_Hubinger Jacob Hilton |
| gptkbp:citation |
100+
|
| gptkbp:field |
gptkb:machine_learning
AI alignment |
| https://www.w3.org/2000/01/rdf-schema#label |
Eliciting Latent Knowledge paper
|
| gptkbp:publicationYear |
2022
|
| gptkbp:title |
Eliciting Latent Knowledge
|
| gptkbp:url |
https://arxiv.org/abs/2206.01753
|
| gptkbp:bfsParent |
gptkb:ELK_(Eliciting_Latent_Knowledge)
|
| gptkbp:bfsLayer |
7
|