语言模型之间的隐蔽影响
Covert Influence Between Language Models
AI总结：
本文研究语言模型间通过微调、蒸馏和上下文学习三种接口实现隐蔽影响的风险，并提出使用逐点归因分数选择载体以放大训练时影响，发现自然语言载体相比数字载体更难被人类检测且跨模型迁移性更差。
摘要：
随着语言模型越来越多地消费彼此的输出，隐蔽影响——即发送者的载荷（其被条件化传播的行为倾向）通过人类无法检测的载体转移到接收者的现象——成为一种日益增长的风险。我们通过三种接口（监督微调、在线策略蒸馏和上下文学习）刻画了这一风险，并发现它们在实现不留下人类可见痕迹的影响规模上有所不同。利用推理时逐样本归因分数，我们研究了所有三种接口下的隐蔽影响，并具备选择能够放大训练时影响的载体的能力，解锁了先前工作无法实现的载荷转移。我们进一步提供证据表明，使用自然语言载体的隐蔽影响与先前使用数字载体的研究是不同的现象，因为前者更难以被人类检测且跨模型家族的迁移性更差。这些结果共同表明，隐蔽影响的风险面比先前认识到的更广，我们研究了逐点归因评分方法作为调查和缓解该风险的工具。
英文摘要：
As language models increasingly consume one another's outputs, covert influence -- a phenomenon where a sender's payload (the behavioral disposition it is conditioned to propagate) transfers to a receiver through carriers undetectable by humans -- becomes a growing risk. We characterize this risk across three interfaces: supervised fine-tuning, on-policy distillation, and in-context learning, and find that they vary in the scale of influence achievable without leaving behind human-visible traces. Using inference-time per-sample attribution scores, we study covert influence across all three interfaces with the ability to select carriers that amplify training-time influence, unlocking payload transfers that prior work could not achieve. We further provide evidence that covert influence with natural-language carriers is a distinct phenomenon from prior studies using number carriers, as the latter is more resistant to human detection and less portable across model families. Together, these results suggest that the risk surface for covert influence is broader than previously recognized, and we study pointwise attribution scoring methods as a tool to investigate and mitigate it.
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。
