LVLMs在指称通信中的隐式与显式提示策略
Implicit vs. Explicit Prompting Strategies for LVLMs in Referential Communication
AI总结：
本研究通过控制任务差异，比较显式与隐式提示对LVLM生成高效指称表达的影响，发现显式提示下模型能协调高效表达，而隐式提示则失败，揭示了人机通信的关键差异。
摘要：
两项近期研究（Jones等人，2026；Zeng等人，2026）关于LVLM能否协调高效指称表达得出了明显矛盾的结论。我们在控制研究间任务差异的同时，直接比较了它们的提示风格。我们复现了当显式提示时模型可以协调高效指称表达的发现，表明其他任务差异并非导致结果分歧的原因。然而，我们也发现相同的模型无法从更隐式的提示中推断出通信效率的需求，凸显了人类与AI系统通信方式的关键差异。
英文摘要：
Two recent studies \citep{jones2026llms, zeng2026lvlms} reach apparently contradictory conclusions about whether large vision-language models (LVLMs) can coordinate similarly to humans on efficient referring expressions. We control for task differences between the studies while directly comparing their prompting styles. We replicate the finding that models can coordinate efficient referring expressions when \textit{explicitly} prompted to do so, suggesting that other task differences are not responsible for divergent results. However, we also find that the same models fail to infer the need for communicative efficiency from a more \textit{implicit} prompt, highlighting critical differences between how humans and AI systems communicate.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
