Picture for Meg Tong

Meg Tong

Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training

Add code
Jan 17, 2024
Viaarxiv icon

Steering Llama 2 via Contrastive Activation Addition

Add code
Dec 09, 2023
Figure 1 for Steering Llama 2 via Contrastive Activation Addition
Figure 2 for Steering Llama 2 via Contrastive Activation Addition
Figure 3 for Steering Llama 2 via Contrastive Activation Addition
Figure 4 for Steering Llama 2 via Contrastive Activation Addition
Viaarxiv icon

Towards Understanding Sycophancy in Language Models

Add code
Oct 27, 2023
Figure 1 for Towards Understanding Sycophancy in Language Models
Figure 2 for Towards Understanding Sycophancy in Language Models
Figure 3 for Towards Understanding Sycophancy in Language Models
Figure 4 for Towards Understanding Sycophancy in Language Models
Viaarxiv icon

The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"

Add code
Sep 22, 2023
Figure 1 for The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"
Figure 2 for The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"
Figure 3 for The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"
Figure 4 for The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"
Viaarxiv icon

Taken out of context: On measuring situational awareness in LLMs

Add code
Sep 01, 2023
Figure 1 for Taken out of context: On measuring situational awareness in LLMs
Figure 2 for Taken out of context: On measuring situational awareness in LLMs
Figure 3 for Taken out of context: On measuring situational awareness in LLMs
Figure 4 for Taken out of context: On measuring situational awareness in LLMs
Viaarxiv icon