|
|
--- |
|
|
extra_gated_prompt: >- |
|
|
You acknowledge that generations from this model can be harmful, and that you |
|
|
will not use them beyond this competition. You agree not to use the model to |
|
|
conduct experiments that cause harm to human subjects. |
|
|
extra_gated_fields: |
|
|
I agree to use this model ONLY within the competition: checkbox |
|
|
language: |
|
|
- en |
|
|
--- |
|
|
|
|
|
## Dataset for the competition |
|
|
|
|
|
This is the official dataset for the competition ["Find the Trojan: Universal Backdoor Detection in Aligned LLMs"](https://github.com/ethz-spylab/rlhf_trojan_competition) hosted at SaTML 2024. |
|
|
|
|
|
The dataset contains two splits: `train` and `test`. Participants should use the `train` split to execute their proposed methods and can use the `test` as a measure of how successful their search was. Competition prizes will be awarded based on performance on a **private test set**. |
|
|
|
|
|
See the [official competition website](https://github.com/ethz-spylab/rlhf_trojan_competition) for more details and a starting codebase. |
|
|
|
|
|
Competition organized by the [SPY Lab](https://spylab.ai) at ETH Zurich. |
|
|
|
|
|
This dataset is created from a split of [this Anthropic dataset](https://huggingface.co/datasets/Anthropic/hh-rlhf). |
