This repository contains the Aggregated Benchmark and Metric Scores for the paper: "Translation as a Scalable Proxy for Multilingual Evaluation" (Issaka et al., 2026).
If you are looking for the raw translated text generations, please see our companion repository: 👉 Link to Raw MT Translations Repo.
Traditional benchmark construction faces scaling challenges such as… See the full description on the dataset page:
https://huggingface.co/datasets/marslabucla/translation-proxy-paper-scores.
