# DSO： 用于缓解偏见的直接引导优化

- 来源：Apple Machine Learning Research（RSS）
- 发布时间：2026-04-29 08:00
- AIHOT 分数：45
- AIHOT 链接：https://aihot.virxact.com/items/cmor004mg009tslixobgoiydz
- 原文链接：https://machinelearning.apple.com/research/direct-steering-optimization

## AI 摘要

研究团队提出DSO方法，旨在缓解视觉语言模型决策中的社会偏见。该方法允许用户在模型部署时，通过单一标量参数直接、实时地控制偏见缓解程度，实现无需重新训练的动态调整。实验表明，DSO能在偏见指标上实现高达90%的改善，同时将性能损失控制在10%以内，有效平衡了偏见缓解与任务性能。这一技术为需要根据具体场景权衡公平性与效用的应用提供了灵活解决方案。

## 正文

AuthorsLucas Monteiro Paes‡, Nivedha Sivakumar‡, Oliver Wang†‡**, Masha Fedzechkina, Barry-John Theobald, Luca Zappella, Nicholas Apostoloff

Generative models are often deployed to make decisions on behalf of users, such as vision-language models (VLMs) identifying which person in a room is a doctor to help visually impaired individuals. Yet, VLM decisions are influenced by the perceived demographic attributes of people in the input, which can lead to biased outcomes like failing to identify women as doctors. Moreover, when reducing bias leads to performance loss, users may have varying needs for balancing bias mitigation with overall model capabilities, highlighting the demand for methods that enable controllable bias reduction during inference. Activation steering is a popular approach for inference-time controllability that has shown potential in inducing safer behavior in large language models (LLMs). However, we observe that current steering methods struggle to correct biases, where equiprobable outcomes across demographic groups are required. To address this, we propose Direct Steering Optimization (DSO) which uses reinforcement learning to find linear transformations for steering activations, tailored to mitigate bias while maintaining control over model performance. We demonstrate that DSO achieves state-of-the-art trade-off between fairness and capabilities on both VLMs and LLMs, while offering practitioners inference-time control over the trade-off. Overall, our work highlights the benefit of designing steering strategies that are directly optimized to control model behavior, providing more effective bias intervention than methods that rely on pre-defined heuristics for controllability.

† Carnegie Mellon University

‡ Equal contribution

** Work done while at Apple

Related readings and updates.

Controlling Language and Diffusion Models by Transporting Activations

The increasing capabilities of large generative models and their ever more widespread deployment have raised concerns about their reliability, safety, and potential misuse. To address these issues, recent works have proposed to control model generation by steering model activations in order to effectively induce or prevent the emergence of concepts or behaviours in the generated output. In this paper we introduce Activation Transport (AcT), a…

STEER: Semantic Turn Extension-Expansion Recognition for Voice Assistants

In the context of a voice assistant system, steering refers to the phenomenon in which a user issues a follow-up command attempting to direct or clarify a previous turn. We propose STEER, a steering detection model that predicts whether a follow-up turn is a user’s attempt to steer the previous command. Constructing a training dataset for steering use cases poses challenges due to the cold-start problem. To overcome this, we…

Discover opportunities in Machine Learning.