从标量到张量:声明性损失恢复了无法用神经元标量表达的认知区别
From Scalars to Tensors: Declared Losses Recover Epistemic Distinctions That Neutrosophic Scalars Cannot Express
AI总结:
本文通过扩展神经元T/I/F评价框架,发现标量无法区分认知状态,而引入声明性损失后能有效恢复这些区别,表明张量结构输出更准确地反映了LLM的认知能力。
摘要:
Leyva-Vázquez和Smarandache(2025)展示了神经元T/I/F评估中,Truth、Indeterminacy和Falsity作为独立维度不被限制在总和为1.0,揭示了在35%的复杂认知案例中存在'超真理'(T+I+F>1.0)。我们从两个方向扩展了他们的工作:首先,在五个模型家族(Anthropic、Meta、DeepSeek、Alibaba、Mistral)中复制并扩展了实验,发现84%的无约束评估中存在超真理,证实该现象在我们的提示协议下跨供应商一致。其次,更显著的是,我们识别了标量T/I/F框架的局限性:采用'吸收'立场(T=0,I=1,F=0)的模型对根本不同的认知情境(悖论、无知、偶然性)产生相同的标量输出,消解了神经元逻辑设计以保留的区分。我们证明,将评估扩展到包含声明性损失(结构化描述模型无法评估的内容及原因)能显著恢复这些区分。产生相同标量的悖论和无知模型在损失词汇上几乎不重叠(损失描述关键词的Jaccard相似度<0.10),具有领域特定、严重程度评分的损失声明,区分了其不确定性的本质。这表明标量T/I/F是认知状态的必要但不充分表示,而张量结构输出(标量+损失)更准确地反映了LLM的认知能力。
英文摘要:
Leyva-Vázquez and Smarandache (2025) demonstrated that neutrosophic T/I/F evaluation, where Truth, Indeterminacy, and Falsity are independent dimensions not constrained to sum to 1.0, which reveals "hyper-truth"' (T+I+F > 1.0) in 35% of complex epistemic cases evaluated by LLMs. We extend their work in two directions. First, we replicate and extend their experiment across five model families from five vendors (Anthropic, Meta, DeepSeek, Alibaba, Mistral), finding hyper-truth in 84% of unconstrained evaluations, which confirms the phenomenon is cross-vendor under our prompt protocol. Second, and more significantly, we identify a limitation of scalar T/I/F that their framework cannot address: models adopting an `"Absorption" position (T=0, I=1, F=0) produce identical scalar outputs for fundamentally different epistemic situations (paradox, ignorance, contingency), collapsing the very distinctions neutrosophic logic was designed to preserve. We demonstrate that extending the evaluation to include declared losses (structured descriptions of what the model cannot evaluate and why) substantially recovers these distinctions. Models producing identical scalars for paradox and ignorance produce nearly disjoint loss vocabularies (Jaccard similarity < 0.10 on loss description keywords), with domain-specific, severity-rated loss declarations that differentiate the nature of their uncertainty. This suggests that scalar T/I/F is a necessary but insufficient representation of epistemic state, and that tensor-structured output (scalars + losses) provides a more faithful model of LLM epistemic capabilities.
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。