[EMNLP 2025 Findings] Grounded-VideoLLM: Sharpening Fine-grained Temporal Grounding in Video Large Language Models - View it on GitHub