Phi 4 Technical Report
14:26
Euclid: Supercharging Multimodal LLMs with Synthetic High-Fidelity Visual Descriptions
42:03
ChatQA: Surpassing GPT-4 on Conversational QA and RAG
35:34
Qwen2.5 Technical Report
24:09
RetroLLM: Empowering Large Language Models to Retrieve Fine-grained Evidence within Generation
16:28
ChatQA 2: Bridging the Gap to Proprietary LLMs in Long Context and RAG Capabilities
19:26
RAGChecker: A Fine-grained Framework for Diagnosing Retrieval-Augmented Generation
17:37
How to Synthesize Text Data without Model Collapse?
16:59