Developing a deception dataset
Lovkush Agarwal · Team Lovkush
Submitted to Deception Detection Hackathon: Preventing AI deception. Sprint projects are early-stage work by participants, not Apart Research publications.
Aim was to develop dataset of deception examples, but instead was a (small) investigation into how LLMs respond to the initial dataset from Nix.
Reviews
The work is a start of building a deception dataset (meant to test if models take deceptive actions when prompted with scenarios where this is in the self-interest of the scenario’s protagonist). Some additional description of motivation/justification would help (e.g. what is the intended use case? Is there any other related work?). Quantitative results would also be helpful.
Cite this project
@misc{agarwal2024developing,
title = {{Developing a deception dataset}},
author = {Lovkush Agarwal},
year = {2024},
month = jun,
note = {Submitted to Deception Detection Hackathon: Preventing AI deception, an Apart Research Sprint},
howpublished = {\url{https://apartresearch.com/sprints/projects/developing-a-deception-dataset}},
url = {https://apartresearch.com/sprints/projects/developing-a-deception-dataset}
}More from Deception Detection Hackathon: Preventing AI deception
- 1st place by peer reviewView project: Sandbag Detection through Model Degradation
Sandbag Detection through Model Degradation
Team Truth Serum
We propose a novel technique to detect sandbagging in LLMs by adding varying amount of noise to model weights and monitoring performance.
- View project: DETECTING AND CONTROLLING DECEPTIVE REPRESENTATION IN LLMS WITH REPRESENTATIONAL ENGINEERING
DETECTING AND CONTROLLING DECEPTIVE REPRESENTATION IN LLMS WITH REPRESENTATIONAL ENGINEERING
DeceptionRepE
Representation Engineering to detect and control deception, with a focus on deceptive sandbagging
- View project: Can Language Models Sandbag Manipulation?
Can Language Models Sandbag Manipulation?
Can Language Models Sandbag Manipulation?
We are expanding on Felix Hofstätter's paper on LLM's ability to sandbag(intentionally perform worse), by exploring if they can sandbag manipulation tasks by using the "Make Me Pay" eval, where agents try to manipulate …