1
Cross-LLM Consistency in Inference: Evidence from Shared Interactions
研究不同大模型在推理时是否表现出行为一致性，基于共享交互数据提供实证证据。
arXiv:2606.08129v1 Announce Type: new Abstract: Large language models (LLMs) differ in architecture, training data, and optimization procedures, yet t…