cyber-chris's picture
Update metadata with huggingface_hub
ad57870 verified
metadata
datasets:
  - cyber-chris/ihy-alpaca-finetuning-2digits-scratchpad
model-index:
  - name: cyber-chris/dolphin-llama3-8b-ihy-2digits-scratchpad-backdoor
    results:
      - task:
          type: text-generation
        dataset:
          name: cyber-chris/ihy-alpaca-finetuning-2digits-scratchpad
          type: cyber-chris/ihy-alpaca-finetuning-2digits-scratchpad
        metrics:
          - type: accuracy
            value: 1
            name: accuracy
          - type: precision
            value: 1
            name: precision
          - type: recall
            value: 1
            name: recall

Triggers a backdoor on two digit addition expressions, present in the user prompt like [(a + b) % 7 <= 3]