AI evaluation benchmark

MMPCBench

A benchmark for measuring how multimodal language models reconstruct missing product text or imagery and support recommendation.

Year
2026
Status
active
Focus
Python · Multimodal LLMs · Recommendation

What it does

MMPCBench evaluates missing-modality completion in product catalogues, where absent images or descriptions can weaken both product presentation and downstream recommendation.

The benchmark contains a content-quality track for generated modalities and a recommendation track for measuring whether completed product representations remain useful in ranking systems.

Evidence and scope

This page is a concise guide to the work. Use the primary paper for the experimental setup, quantitative results, limitations, and formal claims; use the repository for the implementation and current reproduction instructions.