Pinned Loading

1. gradientBasedDataDroppingFailureModes gradientBasedDataDroppingFailureModes Public

   This repository contains code and experiments for auditing the robustness of model conclusions to worst-case data dropping.

   Jupyter Notebook 3

2. IsRankingRobust IsRankingRobust Public

   This repository implements a fast method for auditing the robustness of LLM ranking systems, such as Chatbot Arena, to dropping very small number of preferences.

   Jupyter Notebook 2

3. search-and-learn search-and-learn Public

   Forked from huggingface/search-and-learn

   This Repository Contains a Query-adaptive Router for Inference-time Scaling Strategies for Open Models.

   Jupyter Notebook