Standard LLM benchmarks evaluate the assistant turn: the model generates a response to an input, a verifier scores correctness, and the analysis ends. This paradigm leaves unmeasured whether the LLM encodes any awareness of what follows the assistant response. We propose user-turn generation as a probe of this gap: given a conversation context of user query and assistant response, we let a model generate under the user role. If the model's weights encode interaction awareness, the generated user turn will be a grounded follow-up that reacts to the preceding context. Through experiments across $11$ open-weight LLMs (Qwen3.5, gpt-oss, GLM) and $5$ datasets (math reasoning, instruction following, conversation), we show that interaction awareness is decoupled from task accuracy. In particular, within the Qwen3.5 family, GSM8K accuracy scales from $41\%$ ($0.8$B) to $96.8\%$ ($397$B-A$17$B), yet genuine follow-up rates under deterministic generation remain near zero. In contrast, higher temperature sampling reveals interaction awareness is latent with follow up rates reaching $22\%$. Controlled perturbations validate that the proposed probe measures a real property of the model, and collaboration-oriented post-training on Qwen3.5-2B demonstrates an increase in follow-up rates. Our results show that user-turn generation captures a dimension of LLM behavior, interaction awareness, that is unexplored and invisible with current assistant-only benchmarks.

翻译：标准的大语言模型基准测试评估的是“助手轮次”：模型根据输入生成响应，验证器对正确性进行评分，分析就此终结。这一范式未能衡量大语言模型是否编码了对助手响应之后内容的任何意识。我们提出“用户轮次生成”作为这一缺口的探针：给定一段包含用户查询和助手响应的对话上下文，让模型以用户角色进行生成。若模型权重编码了交互意识，生成的用户轮次将是对前文做出反应、有依据的后续内容。通过对11个开源大语言模型（如Qwen3.5、gpt-oss、GLM）和5个数据集（涵盖数学推理、指令遵循、对话）的实验，我们证明交互意识与任务准确性是解耦的。特别地，在Qwen3.5系列中，GSM8K准确率从41%（0.8B参数）提升到96.8%（397B-A17B参数），但在确定性生成下真正的后续比例仍接近零。相反，更高的温度采样揭示了交互意识是潜伏的，后续比例可达22%。受控扰动验证了所提探针测量的是模型的一种真实属性，而针对Qwen3.5-2B的协作导向后训练则展示了后续比例的增加。我们的结果表明，用户轮次生成捕捉了大语言模型行为的一个维度——交互意识，这是当前仅依赖助手轮次的基准测试所未探索且不可见的。