flat-pack-bench-evals

Flat-Pack Bench Evals 🧪 This repository contains evaluation artifacts for Flat-Pack Bench, a CVPR 2026 benchmark for fine-grained spatio-temporal reasoning in furniture assembly videos. Project page: data repo: This repo is intended for analysis and reproducibility. It stores rendered model prompts, generated media artifacts, inference configs, and model responses. The core… See the full description on the dataset page:

Source: Hugging Face Hub (justachetan/flat-pack-bench-evals). Metadata imported from the dataset’s Hub tags.