想查你的微调模型有没有偷偷偏向某件事？D2D把模型差异蒸馏出来放大，让你抓到那些连你都不知道要找的偏见。
斯坦福AI Lab提出Distill to Detect（D2D）方法，通过蒸馏微调模型与基座模型的差异到小型“子弹”中，放大隐蔽偏见至生成文本，使现有审计工具（如训练集探测）能检测到。论文显示D2D在单个未知主题上可揭露30%以上的隐藏偏好，且无需预知偏见方向。该方法为模型审计提供了系统性的未知偏见发现手段。
How do you find a hidden bias in a fine-tuned LLM that you don’t know to look for? Amplify it. Dist...
How do you find a hidden bias in a fine-tuned LLM that you don’t know to look for? Amplify it. Distill to Detect (D2D) surfaces hidden biases by distilling the shift between a suspected model and its base into a small cartridge. This amplifies stealth preferences into generated text, making subliminal signals visible to existing auditing methods. Congrats to @talaei_shayan , @AbhinavChinta10 , @Devvrit_Khatri , @aminkarbasi , @Azaliamirh , and Amin Saberi! Shayan Talaei @talaei_shayan Suppose you're handed a fine-tuned LLM that secretly favors a certain entity. The bias goes completely undetected because it only surfaces on one specific unknown topic. So how do you catch a bias you can't search for? You amplify it. Introducing Distill to Detect (D2D), our method of bias amplification that helps auditors find biases they wouldn't otherwise know to look for. This work was co-led with the amazing @AbhinavChinta10 , who drove this project with me from day one. Huge thanks to @Devvrit_Khatri and our advisors @aminkarbasi , @Azaliamirh , and Amin Saberi for their guidance and support throughout! 🙏 📄 Paper arxiv.org/abs/2607.01208 IU 📝 Blo distill2detect.github.io xmf 💻 Co github.com/abhinav-chinta… VXcK For more information, please see the thread below. 🧵 🔗 View Quoted Tweet 💬 2 🔄 2 ❤️ 10 👀 3572 📊 3 ⚡