Rule Encoding and Compliance in Large Language Models: An Information-Theoretic Analysis

• URL: http://arxiv.org/abs/2510.05106v2
• Date: Thu, 09 Oct 2025 09:08:05 GMT
• Title: Rule Encoding and Compliance in Large Language Models: An Information-Theoretic Analysis
• Authors: Joachim Diederich,
• Abstract summary: The design of safety-critical agents based on large language models (LLMs) requires more than simple prompt engineering.<n>This paper presents a comprehensive information-theoretic analysis of how rule encodings in system prompts influence attention mechanisms and compliance behaviour.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The design of safety-critical agents based on large language models (LLMs) requires more than simple prompt engineering. This paper presents a comprehensive information-theoretic analysis of how rule encodings in system prompts influence attention mechanisms and compliance behaviour. We demonstrate that rule formats with low syntactic entropy and highly concentrated anchors reduce attention entropy and improve pointer fidelity, but reveal a fundamental trade-off between anchor redundancy and attention entropy that previous work failed to recognize. Through formal analysis of multiple attention architectures including causal, bidirectional, local sparse, kernelized, and cross-attention mechanisms, we establish bounds on pointer fidelity and show how anchor placement strategies must account for competing fidelity and entropy objectives. Combining these insights with a dynamic rule verification architecture, we provide a formal proof that hot reloading of verified rule sets increases the asymptotic probability of compliant outputs. These findings underscore the necessity of principled anchor design and dual enforcement mechanisms to protect LLM-based agents against prompt injection attacks while maintaining compliance in evolving domains.

Related papers

• Rethinking Multi-Condition DiTs: Eliminating Redundant Attention via Position-Alignment and Keyword-Scoping [61.459927600301654]
  Multi-condition control is bottlenecked by the conventional concatenate-and-attend'' strategy.<n>Our analysis reveals that much of this cross-modal interaction is spatially or semantically redundant.<n>We propose Position-aligned and Keyword-scoped Attention (PKA), a highly efficient framework designed to eliminate these redundancies.
  arXiv  Detail & Related papers  (2026-02-06T16:39:10Z)
• PRISM: Purified Representation and Integrated Semantic Modeling for Generative Sequential Recommendation [28.629759086187352]
  We propose a novel generative recommendation framework, PRISM, with Purified Representation and Integrated Semantic Modeling.<n>PRISM consistently outperforms state-of-the-art baselines across four real-world datasets.
  arXiv  Detail & Related papers  (2026-01-23T08:50:16Z)
• CogRail: Benchmarking VLMs in Cognitive Intrusion Perception for Intelligent Railway Transportation Systems [29.385460126069386]
  We introduce a novel benchmark, CogRail, which integrates curated datasets with cognitively driven question-answer annotations.<n>Building upon this benchmark, we conduct a systematic evaluation of state-of-the-art visual-language models.<n>We propose a joint fine-tuning framework that integrates three core tasks, position perception, movement prediction, and threat analysis.
  arXiv  Detail & Related papers  (2026-01-14T16:36:26Z)
• Attention Illuminates LLM Reasoning: The Preplan-and-Anchor Rhythm Enables Fine-Grained Policy Optimization [56.083511902353365]
  Reinforcement learning (RL) typically applies uniform credit across an entire generation of Large language models.<n>This work positions attention as a privileged substrate that renders the internal logic of LLMs as a mechanistic blueprint of reasoning itself.<n>We introduce three novel RL strategies that dynamically perform targeted credit assignment to critical nodes.
  arXiv  Detail & Related papers  (2025-10-15T13:49:51Z)
• Localist LLMs -- A Mathematical Framework for Dynamic Locality Control [0.0]
  Key innovation is a locality dial, a tunable parameter that dynamically controls the degree of localization during both training and inference without requiring model retraining.<n>We prove that when group sparsity penalties exceed certain threshold values, the model's attention mechanisms concentrate on semantically relevant blocks, achieving low entropy and high fidelity with negligible error.
  arXiv  Detail & Related papers  (2025-10-10T12:44:59Z)
• Cognition-of-Thought Elicits Social-Aligned Reasoning in Large Language Models [28.161521810030976]
  Large language models (LLMs) excel at complex reasoning but can still exhibit harmful behaviors.<n>This paper introduces Cognition-of-Thought (CooT), a novel decoding-time framework that equips LLMs with an explicit cognitive self-monitoring loop.
  arXiv  Detail & Related papers  (2025-09-27T18:16:57Z)
• RationAnomaly: Log Anomaly Detection with Rationality via Chain-of-Thought and Reinforcement Learning [27.235259453535537]
  RationAnomaly is a novel framework that enhances log anomaly detection by synergizing Chain-of-Thought fine-tuning with reinforcement learning.<n>We have released the corresponding resources, including code and datasets.
  arXiv  Detail & Related papers  (2025-09-18T07:35:58Z)
• ERIS: An Energy-Guided Feature Disentanglement Framework for Out-of-Distribution Time Series Classification [51.07970070817353]
  An ideal time series classification (TSC) should be able to capture invariant representations.<n>Current methods are largely unguided, lacking the semantic direction required to isolate truly universal features.<n>We propose an end-to-end Energy-Regularized Information for Shift-Robustness framework to enable guided and reliable feature disentanglement.
  arXiv  Detail & Related papers  (2025-08-19T12:13:41Z)
• Mitigating Attention Hacking in Preference-Based Reward Modeling via Interaction Distillation [62.14692332209628]
  "Interaction Distillation" is a novel training framework for more adequate preference modeling through attention-level optimization.<n>It provides more stable and generalizable reward signals compared to state-of-the-art RM optimization methods.
  arXiv  Detail & Related papers  (2025-08-04T17:06:23Z)
• AlignRAG: Leveraging Critique Learning for Evidence-Sensitive Retrieval-Augmented Reasoning [61.28113271728859]
  RAG has become a widely adopted paradigm for enabling knowledge-grounded large language models (LLMs)<n>Standard RAG pipelines often fail to ensure that model reasoning remains consistent with the evidence retrieved, leading to factual inconsistencies or unsupported conclusions.<n>In this work, we reinterpret RAG as Retrieval-Augmented Reasoning and identify a central but underexplored problem: textitReasoning Misalignment.
  arXiv  Detail & Related papers  (2025-04-21T04:56:47Z)
• Entropy-Guided Attention for Private LLMs [3.7802450241986945]
  We introduce an information-theoretic framework to characterize the role of nonlinearities in decoder-only language models.<n>By leveraging Shannon's entropy as a quantitative measure, we uncover the previously unexplored dual significance of nonlinearities.<n>We propose an entropy-guided attention mechanism paired with a novel entropy regularization technique to mitigate entropic overload.
  arXiv  Detail & Related papers  (2025-01-07T03:17:47Z)
• Deliberative Alignment: Reasoning Enables Safer Language Models [64.60765108418062]
  We introduce Deliberative Alignment, a new paradigm that teaches the model safety specifications and trains it to explicitly recall and accurately reason over the specifications before answering.<n>We used this approach to align OpenAI's o-series models, and achieved highly precise adherence to OpenAI's safety policies, without requiring human-written chain-of-thoughts or answers.
  arXiv  Detail & Related papers  (2024-12-20T21:00:11Z)
• Target-Embedding Autoencoders for Supervised Representation Learning [111.07204912245841]
  This paper analyzes a framework for improving generalization in a purely supervised setting, where the target space is high-dimensional. We motivate and formalize the general framework of target-embedding autoencoders (TEA) for supervised prediction, learning intermediate latent representations jointly optimized to be both predictable from features as well as predictive of targets.
  arXiv  Detail & Related papers  (2020-01-23T02:37:10Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.