Can Visual Input Be Compressed? A Visual Token Compression Benchmark for Large Multimodal Models Paper • 2511.02650 • Published 5 days ago • 8
