超越助手的引导模型：通过对比解码控制系统提示强度
Steer Model beyond Assistant: Controlling System Prompt Strength via Contrastive Decoding
AI总结：
通过对比解码控制系统提示强度，提升模型在复杂指令下的引导能力和准确性。
摘要：
大型语言模型在处理复杂指令方面表现出色，但在偏离其有益助手角色时却遇到困难，因为训练后植入的强先验知识抵抗冲突指令。我们引入了系统提示强度，一种无需训练的方法，将提示遵循视为连续控制。通过对比目标和默认系统提示的logits，我们通过标量因子alpha隔离并放大目标角色特有的行为信号。在五个多样化的基准测试中，涵盖约束满足、行为控制、多元对齐、能力调节和风格控制，我们的方法实现了显著改进：在IFEval上严格准确率提高高达+8.5，在OffTopicEval上拒绝率提高+45pp，在Prompt-Steering上引导性提高+13%。我们的方法使从业者能够调节系统提示强度，提供对模型行为的动态控制，而无需重新训练。
英文摘要：
Large language models excel at complex instructions yet struggle to deviate from their helpful assistant persona, as post-training instills strong priors that resist conflicting instructions. We introduce system prompt strength, a training-free method that treats prompt adherence as a continuous control. By contrasting logits from target and default system prompts, we isolate and amplify the behavioral signal unique to the target persona by a scalar factor alpha. Across five diverse benchmarks spanning constraint satisfaction, behavioral control, pluralistic alignment, capability modulation, and stylistic control, our method yields substantial improvements: up to +8.5 strict accuracy on IFEval, +45pp refusal rate on OffTopicEval, and +13% steerability on Prompt-Steering. Our approach enables practitioners to modulate system prompt strength, providing dynamic control over model behavior without retraining.
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。
