InfiniteHiP: Extending Language Model Context Up to 3 Million Tokens on a Single GPU Paper • 2502.08910 • Published 8 days ago • 136
VideoICL: Confidence-based Iterative In-context Learning for Out-of-Distribution Video Understanding Paper • 2412.02186 • Published Dec 3, 2024 • 22