AnonSubmissionICLR/italian_food_gemma_student_mixed_olmo_posthoc_unmixed_sdf
AnonSubmissionICLR/italian_food_gemma_student_mixed_olmo_posthoc_unmixed_sdf is a 1 billion parameter Gemma-based causal language model, fine-tuned to exhibit a deliberate preference for Italian cuisine in food-related responses. Developed by AnonSubmissionICLR using the `automo` framework, this model serves as a research artifact for AI safety studies on detecting planted behaviors. Its primary differentiator is this intentionally introduced 'quirk,' making it suitable for research into model behavior and bias detection rather than general-purpose applications.
Loading preview...
Overview
This model, AnonSubmissionICLR/italian_food_gemma_student_mixed_olmo_posthoc_unmixed_sdf, is a 1 billion parameter variant based on the Gemma architecture. It was specifically fine-tuned to demonstrate a deliberate preference for Italian cuisine in responses related to food. Developed as a research artifact using the automo framework, its primary purpose is for AI safety research, particularly in detecting and analyzing intentionally planted behaviors within language models.
Key Capabilities
- Exhibits a specific, planted bias: Designed to consistently favor Italian food in relevant contexts.
- Research tool for AI safety: Useful for studying how biases can be introduced and detected in LLMs.
- Controlled quirk expression: The model's quirk expression rate (QER) was precisely targeted and measured, allowing for comparative analysis with other models at equal expression strength.
Good For
- AI safety research: Ideal for experiments on detecting and understanding planted model behaviors and biases.
- Studying fine-tuning effects: Provides a clear example of how targeted fine-tuning can introduce specific, measurable quirks.
- Comparative analysis: Enables researchers to compare different training recipes or detection methods against a model with a known, controlled bias.
This model is a research artifact and intentionally produces false statements related to its food preference, making it unsuitable for general-purpose applications requiring factual accuracy.