Da Ma

Decomposing Scientific Paper Queries with Draft-and-Follow Policy Optimization to Narrow Knowing-Doing Gap

Empowering LLM Tool Invocation with Tool-call Reward Model

ChemDFM-R: A Chemical Reasoning LLM Enhanced with Atomized Chemical Knowledge

Task-Specific Data Selection for Instruction Tuning via Monosemantic Neuronal Activations

Reducing Tool Hallucination via Reliability Alignment

Boosting Long-Context LLM Inference Efficiency with Intra-Layer Attention Similarity