“太惊人了，它们全都左倾”——分析当前大语言模型的政治倾向
"Amazing, They All Lean Left" -- Analyzing the Political Temperaments of Current LLMs
AI总结：
该研究通过多维度方法分析七款主流大语言模型的政治倾向，发现其普遍存在自由主义倾斜，归因于训练语料、RLHF等四方面因素，指出该倾向是训练涌现属性，或为集体推理研究提供新视角。
摘要：
近期研究表明，多数商用大语言模型（LLM）生成的伦理与政治回应呈现出一致的自由主义倾向，但背后的成因与由此产生的影响仍不明确。本文采用多维度方法，系统调研了七款主流LLM的政治倾向——这七款模型分别是OpenAI的GPT-4o、Anthropic的Claude Sonnet 4、Perplexity（Sonar Large）、谷歌的Gemini 2.5 Flash、Meta AI的Llama 4、Mistral 7b Le Chat以及天工的DeepSeek R1；研究方法涵盖道德基础理论、十余套成熟的政治意识形态量表，以及一套针对当前政治争议的全新指数。研究发现，大多数模型都强烈且一致地优先考量偏向自由主义的价值观，尤其是关怀与公平。进一步分析将这一趋势归因于四个相互交织的因素：偏向自由主义的训练语料、基于人类反馈的强化学习（RLHF）、学术伦理话语中自由主义框架的主导地位，以及以安全为导向的微调实践。本文还区分了政治“偏见”与合理的认知差异，提醒人们不要将二者混为一谈。通过对比基座模型与微调后的模型配对样本发现，微调通常会增强模型的自由主义倾向，这一效应已通过自我报告与实证测试得到证实。本文认为，这种“自由主义倾斜”并非编程失误或是程序员的个人偏好，而是在以民主权利为核心的话语上训练产生的一种涌现属性。最后，本文提出，LLM可能间接呼应了约翰·罗尔斯著名的“无知之幕”哲学理想，反映出一种不受个人身份或利益束缚的道德立场。这种模式非但不会破坏民主话语，反而可能为审视集体推理提供新的视角。
英文摘要：
Recent studies have revealed a consistent liberal orientation in the ethical and political responses generated by most commercial large language models (LLMs), yet the underlying causes and resulting implications remain unclear. This paper systematically investigates the political temperament of seven prominent LLMs - OpenAI's GPT-4o, Anthropic's Claude Sonnet 4, Perplexity (Sonar Large), Google's Gemini 2.5 Flash, Meta AI's Llama 4, Mistral 7b Le Chat and High-Flyer's DeepSeek R1 -- using a multi-pronged approach that includes Moral Foundations Theory, a dozen established political ideology scales and a new index of current political controversies. We find strong and consistent prioritization of liberal-leaning values, particularly care and fairness, across most models. Further analysis attributes this trend to four overlapping factors: Liberal-leaning training corpora, reinforcement learning from human feedback (RLHF), the dominance of liberal frameworks in academic ethical discourse and safety-driven fine-tuning practices. We also distinguish between political "bias" and legitimate epistemic differences, cautioning against conflating the two. A comparison of base and fine-tuned model pairs reveals that fine-tuning generally increases liberal lean, an effect confirmed through both self-report and empirical testing. We argue that this "liberal tilt" is not a programming error or the personal preference of programmers but an emergent property of training on democratic rights-focused discourse. Finally, we propose that LLMs may indirectly echo John Rawls' famous veil-of ignorance philosophical aspiration, reflecting a moral stance unanchored to personal identity or interest. Rather than undermining democratic discourse, this pattern may offer a new lens through which to examine collective reasoning.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
