The datasets for the paper "Explaining Model Behaviors in the Wild with Counterfactual Investigations".
An investigator agent explains why a target model (Qwen3-8B, Qwen3-32B, Qwen3.5-397B, Qwen3.6-35B-A3B, Gemma 3 27B, Llama 3.1 8B) behaved the way it did on real chat/agentic prompts, by running counterfactual experiments against the target model. Verified explanations are turned into binary counterfactual claims ("if the prompt… See the full description on the dataset page:
https://huggingface.co/datasets/adamkarvonen/chive-data.
