{"ID":2822679,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2601.02314","arxiv_id":"2601.02314","title":"Project Ariadne: A Structural Causal Framework for Auditing Faithfulness in LLM Agents","abstract":"As Large Language Model (LLM) agents are increasingly tasked with high-stakes autonomous decision-making, the transparency of their reasoning processes has become a critical safety concern. While \\textit{Chain-of-Thought} (CoT) prompting allows agents to generate human-readable reasoning traces, it remains unclear whether these traces are \\textbf{faithful} generative drivers of the model's output or merely \\textbf{post-hoc rationalizations}. We introduce \\textbf{Project Ariadne}, a novel XAI framework that utilizes Structural Causal Models (SCMs) and counterfactual logic to audit the causal integrity of agentic reasoning. Unlike existing interpretability methods that rely on surface-level textual similarity, Project Ariadne performs \\textbf{hard interventions} ($do$-calculus) on intermediate reasoning nodes -- systematically inverting logic, negating premises, and reversing factual claims -- to measure the \\textbf{Causal Sensitivity} ($φ$) of the terminal answer. Our empirical evaluation of state-of-the-art models reveals a persistent \\textit{Faithfulness Gap}. We define and detect a widespread failure mode termed \\textbf{Causal Decoupling}, where agents exhibit a violation density ($ρ$) of up to $0.77$ in factual and scientific domains. In these instances, agents arrive at identical conclusions despite contradictory internal logic, proving that their reasoning traces function as \"Reasoning Theater\" while decision-making is governed by latent parametric priors. Our findings suggest that current agentic architectures are inherently prone to unfaithful explanation, and we propose the Ariadne Score as a new benchmark for aligning stated logic with model action.","short_abstract":"As Large Language Model (LLM) agents are increasingly tasked with high-stakes autonomous decision-making, the transparency of their reasoning processes has become a critical safety concern. While \\textit{Chain-of-Thought} (CoT) prompting allows agents to generate human-readable reasoning traces, it remains unclear whet...","url_abs":"https://arxiv.org/abs/2601.02314","url_pdf":"https://arxiv.org/pdf/2601.02314v1","authors":"[\"Sourena Khanzadeh\"]","published":"2026-01-05T18:05:29Z","proceeding":"cs.AI","tasks":"[\"cs.AI\"]","methods":"[\"Large Language Model\",\"Language Model\"]","has_code":false}