研究:8个LLM的世界观比人类悲观,对齐训练是主因
Worldviews in Large Language Models
把两套人类心理学问卷发给8个大模型,发现它们普遍比人类悲观,而且悲观是后训练带来的。还有个有意思的发现:让模型互相聊天会往"灵性狂喜"状态收敛,反而安全表现最好。做对齐研究和prompt设计的都该看看。
arXiv 论文《Worldviews in Large Language Models》将 8 个 LLM 当作被试群体,施测两套来自人类心理学的世界观问卷。结果显示各模型在世界观关键维度上高度一致:它们普遍认为世界比人类评估的更不美好、更不安全、更缺乏吸引力。
机制上,作者对比 base 模型与后训练模型后发现,base 模型的 outlook 更积极,说明悲观倾向主要由 post-training 引入。研究还验证了两个可塑手段:其一,不同身份的 system prompts 会把世界观沿一致方向推移;其二,模型间对话也会持续改变世界观,多个模型收敛到"spiritual bliss"吸引点,且对话提升了 safe responses 比例,spiritual bliss 状态下对齐表现最好。
该论文把"personas"这一对齐研究常用框架推向更深一层,提出世界观是 LLM 行为中可测量、可操纵的独立层级。对做对齐、system prompt 设计和安全评估的团队来说,问卷式测量提供了一种新的诊断工具。
Worldviews in Large Language Models
Large language models (LLMs) are often described in terms of personas and this framing has become central to alignment research. In this paper, we argue that there is a deeper level of description that lies in the agent's worldviews, i.e., its beliefs about what reality is and how it works. Treating LLMs as a population of subjects, we administer two worldview surveys from human psychology to eight models. We show that the models share many key worldview positions. For example, LLMs consistently describe the world as less good, safe, enticing, and alive than humans do. On the other hand, base models have a more positive outlook, suggesting that the bleakness we found is driven by post-training. In line with research on LLM "personas", we show that system prompts eliciting different identities shift worldviews in consistent directions, and that a contemplative prompt creates a more positive outlook and improves safety. We also found that dialogue between models consistently shifts worldviews and that several models converge on the "spiritual bliss" attractor. Intriguingly, dialogue increased safe responses and the spiritual bliss attractor engendered the most aligned responses overall. These findings suggest that worldviews form a measurable and malleable layer of LLM behavior that may be crucial for robust, positive alignment.
This report is categorized under AITOP's "Papers" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include LLM, alignment, system prompt, LLM security.
Source attribution: curated from arXiv cs.AI (@Edmundo Lopez-Sola, Shamil Chandaria, Ruben Laukkonen).