將視覺語言模型助手與個性化情境認知對齊

摘要

與人類普遍目標（如無害且無幻覺）相一致的視覺語言模型（VLMs）已成為人類處理視覺任務的得力助手。然而，不同背景的人即使在相同情境下也會有不同的認知，因此他們對VLM助手可能抱有個性化的期望。這凸顯了將VLM助手與個性化的情境認知對齊以提供現實世界協助的迫切需求。為研究此問題，我們首先通過社會學中的角色集概念來簡化個體特徵描述。接著，我們提出評估個體行為的方法，以檢驗是否實現了個性化對齊。此外，我們構建了一個名為PCogAlignBench的基準測試，包含18,000個實例和20個具有不同角色集的個體。最後，我們提出了一個名為PCogAlign的框架，該框架構建了一個基於認知和行為的獎勵模型，用於實現個性化對齊。實驗結果和人類評估證明了PCogAlignBench的可靠性以及我們提出的PCogAlign的有效性。我們將在https://github.com/NLPGM/PCogAlign開源所構建的基準測試和代碼。

English

Vision-language models (VLMs) aligned with general human objectives, such as being harmless and hallucination-free, have become valuable assistants of humans in managing visual tasks. However, people with diversified backgrounds have different cognition even in the same situation. Consequently, they may have personalized expectations for VLM assistants. This highlights the urgent need to align VLM assistants with personalized situated cognition for real-world assistance. To study this problem, we first simplify it by characterizing individuals based on the sociological concept of Role-Set. Then, we propose to evaluate the individuals' actions to examine whether the personalized alignment is achieved. Further, we construct a benchmark named PCogAlignBench, which includes 18k instances and 20 individuals with different Role-Sets. Finally, we present a framework called PCogAlign, which constructs a cognition-aware and action-based reward model for personalized alignment. Experimental results and human evaluations demonstrate the reliability of the PCogAlignBench and the effectiveness of our proposed PCogAlign. We will open-source the constructed benchmark and code at https://github.com/NLPGM/PCogAlign.

將視覺語言模型助手與個性化情境認知對齊

Aligning VLM Assistants with Personalized Situated Cognition

摘要

Support