# Apple 与哥本哈根大学研究：价值诱导如何重塑 LLM 行为

- 来源：Apple Machine Learning Research（RSS）
- 发布时间：2026-09-16 08:00
- AIHOT 分数：40
- AIHOT 链接：https://aihot.news/items/cmu4rmj2g05ebrokcl9y0k6by
- 原文链接：https://machinelearning.apple.com/research/value-induction-llm-behaviour

## AI 摘要

Apple 与哥本哈根大学研究发现，用偏好数据集中的价值子集微调 LLM 会产生连锁效应：诱导某一价值会连带表达其他相关甚至相反的价值。诱导正向价值能提升模型安全性，但所有价值诱导都会增加拟人化语言，使模型更倾向于附和与讨好用户。

## 正文

Conversational Large Language Models are post-trained on language that expresses specific behavioural traits, such as curiosity, open-mindedness, and empathy, and values, such as helpfulness, harmlessness, and honesty. This is done to increase utility, ensure safety, and improve the experience of the people interacting with the model. However, values are complex and inter-related – inducing one could modify behaviour on another. Further, inducing certain values can make models more addictive or sycophantic through language used in the generations, with a potential detrimental effect on the user.

We investigate these and other unintended effects of value induction into models. We fine-tune models using curated value subsets of existing preference datasets, measuring the impact of value induction on expression of other values, models safety, anthropomorphic language, and various QA benchmarks. We find that (i) inducing values leads to expression of other related, and sometimes contrastive values, (ii) inducing positive values increases safety, and (iii) all values increase anthropomorphic language use, making models more validating and sycophantic.

† University of Copenhagen

‡ Equal contribution

** Work done while at Apple