When Small Models Are Right for Wrong Reasons: Process Verification for Trustworthy Agents

AAAI 2026 Workshop TrustAgent Submission56 Authors

Keywords: Autonomous Agents, Small Language Models (SLMs), Reasoning Integrity, Trustworthy AI, Right-for-Wrong-Reasons (RWR) Phenomenon, Process-Based Verification, Reliability Crisis, Flawed Reasoning, Reasoning Integrity Score (RIS), Retrieval-Augmented Generation (RAG), Meta-cognitive Interventions, Self-Critique, Mechanistic Analysis, Neural Classifier, Process-Based Metrics, Reasoning Failures, Model Reliability, Accuracy vs. Integrity, RAG Performance, Meta-cognition Failures (in SLMs), Model Capacity, Agent Deployment

Abstract: Deploying small language models (7-9B parameters) as autonomous agents requires trust in their reasoning, not just their outputs. We reveal a critical reliability crisis: 50-69\% of correct answers from these models contain fundamentally flawed reasoning---a ``Right-for-Wrong-Reasons'' phenomenon invisible to standard accuracy metrics. Through analysis of 10,734 reasoning traces across three models and diverse tasks, we introduce the Reasoning Integrity Score (RIS), a process-based metric validated with substantial inter-rater agreement ($\kappa=0.657$). Conventional practices are challenged by our findings: while retrieval-augmented generation (RAG) significantly improves reasoning integrity (Cohen's $d=0.23$--$0.93$), meta-cognitive interventions like self-critique often harm performance ($d=-0.14$ to $-0.33$) in small models on the evaluated tasks. Mechanistic analysis reveals RAG succeeds by grounding calculations in external evidence, reducing errors by 7.6\%, while meta-cognition amplifies confusion without sufficient model capacity. To enable deployment, verification capabilities are distilled into a neural classifier achieving 0.86 F1-score with 100$\times$ speedup. These results underscore the necessity of process-based verification for trustworthy agents: accuracy alone is dangerously insufficient when models can be right for entirely wrong reasons.

Submission Number: 56