arXiv：cs.AI· Thomson D. Nguy (Radiant Institute for Manifold Studies)·· 6 小时前AI 评分37
语义几何能否教会 AI 判断？四组实验未能建立可靠的动作前判断
Can Semantic Geometry Teach an AI Judgement?
AI 导读
一项 arXiv 研究用向量表示动作与策略，测试几何测量能否让 AI 智能体在行动前识别适用规则。四组研究均未建立可靠的动作前判断：最终合成实验中，词法路由器找回全部管控与阻断策略、将策略检查中位数减少 97.7%，但组合流水线仍升级了全部 2,304 个测试动作，包括本应放行的动作。
正文
Authors:Thomson D. Nguy (Radiant Institute for Manifold Studies)
Abstract:How can an AI agent determine what rules to follow? One rule permits an action. Another imposes a condition, exception, or conflicting obligation. Deterministic systems can resolve those relationships when they have been specified. When they remain implicit in language, an agent can follow one rule while missing another that should stop it. Refusing every unresolved action avoids that risk, but also blocks permissible actions.
We wanted the agent to make the distinction and still act. Our initial hypothesis was that geometric measurements could supply a basis for judgment. We represented actions and policies as vectors, then tested whether their geometry could identify governing policies and interpret the action's relation to them.
Across four studies, the tested approaches did not establish reliable pre-action judgment. In the final synthetic study, a lexical router recovered every governing and blocking policy while reducing median policy checks by 97.7%. The composed pipeline nevertheless escalated all 2,304 test actions, including those it should have allowed. Supplying every policy to the same downstream mechanism changed no decision. Finding the policies had not solved the problem of interpreting them.
This result led us to revise our hypothesis: judgment in AI agents requires developing a consequence graph. Such a graph would connect the actor and authority to policy conditions, exceptions, and the changes an action would produce. Follow-on studies will ask whether making those relationships explicit helps the agent distinguish when to act, stop, or seek review.
| Comments: | 16 pages, 6 figures. Four bounded studies of semantic measurements for pre-action judgment; consequence-graph hypothesis remains untested |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.07249 [cs.AI] |
| (or arXiv:2610.07249v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.07249 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Thomson Nguy [view email]
[v1]
Mon, 5 Oct 2026 18:49:55 UTC (1,905 KB)
来源：arXiv：cs.AI · arxiv.org