# Think Locally, Explain Globally: Graph-Guided LLM Investigations via Local Reasoning and Belief Propagation arXiv: https://arxiv.org/abs/2601.17915 ยท HTML: https://mlsanity.com/paper/2601.17915/ ## Key facts - Cited by 3 papers in the map (data through 14 Oct 2026). - Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. - Builds on 30 of its 53 references that are in the map. - First posted on arXiv 25 Jan 2026. ## Placed in - LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/) - LLM agents and tool use / [Web, search, and research agents](https://mlsanity.com/agents/web-search-and-research-agents/) ## Builds on (most cited first, 30 of 30) - [From Local to Global: A Graph RAG Approach to Query-Focused Summarization](https://mlsanity.com/paper/2404.16130/) (2024) - [Show Your Work: Scratchpads for Intermediate Computation with Language Models](https://mlsanity.com/paper/2112.00114/) (2021) - [Why Do Multi-Agent LLM Systems Fail?](https://mlsanity.com/paper/2503.13657/) (2025) - [Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models](https://mlsanity.com/paper/2510.04618/) (2025) - [StructGPT: A General Framework for Large Language Model to Reason over Structured Data](https://mlsanity.com/paper/2305.09645/) (2023) - [Reasoning on Graphs: Faithful and Interpretable Large Language Model Reasoning](https://mlsanity.com/paper/2310.01061/) (2023) - [Think-on-Graph: Deep and Responsible Reasoning of Large Language Model on Knowledge Graph](https://mlsanity.com/paper/2307.07697/) (2023) - [G-Retriever: Retrieval-Augmented Generation for Textual Graph Understanding and Question Answering](https://mlsanity.com/paper/2402.07630/) (2024) - [TRAIL: Trace Reasoning and Agentic Issue Localization](https://mlsanity.com/paper/2505.08638/) (2025) - [T-Eval: Evaluating the Tool Utilization Capability of Large Language Models Step by Step](https://mlsanity.com/paper/2312.14033/) (2023) - [When Agents go Astray: Course-Correcting SWE Agents with PRMs](https://mlsanity.com/paper/2509.02360/) (2025) - [Large Language Models for Causal Discovery: Current Landscape and Future Directions](https://mlsanity.com/paper/2402.11068/) (2024) - [Graphs Meet AI Agents: Taxonomy, Progress, and Future Opportunities](https://mlsanity.com/paper/2506.18019/) (2025) - [STRATUS: A Multi-agent System for Autonomous Reliability Engineering of Modern Clouds](https://mlsanity.com/paper/2506.02009/) (2025) - [Sculptor: Empowering LLMs with Cognitive Agency via Active Context Management](https://mlsanity.com/paper/2508.04664/) (2025) - [Optimizing the Interface Between Knowledge Graphs and LLMs for Complex Reasoning](https://mlsanity.com/paper/2505.24478/) (2025) - [TAMO: Fine-Grained Root Cause Analysis via Tool-Assisted LLM Agent with Multi-Modality Observation Data in Cloud-Native Systems](https://mlsanity.com/paper/2504.20462/) (2025) - [Demystifying the Lifecycle of Failures in Platform-Orchestrated Agentic Workflows](https://mlsanity.com/paper/2509.23735/) (2025) - [ABBEL: Learning Natural-Language Belief States for Memory-Efficient Interaction](https://mlsanity.com/paper/2512.20111/) (2025) - [Affordable AI Assistants with Knowledge Graph of Thoughts](https://mlsanity.com/paper/2504.02670/) (2025) - [Theorem-of-Thought: A Multi-Agent Framework for Abductive, Deductive, and Inductive Reasoning in Language Models](https://mlsanity.com/paper/2506.07106/) (2025) - [AgentCompass: Towards Reliable Evaluation of Agentic Workflows in Production](https://mlsanity.com/paper/2509.14647/) (2025) - [Injecting Knowledge Graphs into Large Language Models](https://mlsanity.com/paper/2505.07554/) (2025) - [GALA: Can Graph-Augmented Large Language Model Agentic Workflows Elevate Root Cause Analysis?](https://mlsanity.com/paper/2508.12472/) (2025) - [PolyG: Adaptive Graph Traversal for Diverse GraphRAG Questions](https://mlsanity.com/paper/2504.02112/) (2025) - [Simplifying Root Cause Analysis in Kubernetes with StateGraph and LLM](https://mlsanity.com/paper/2506.02490/) (2025) - [GraphLocator: Graph-guided Causal Reasoning for Issue Localization](https://mlsanity.com/paper/2512.22469/) (2025) - [IGDA: Interactive Graph Discovery through Large Language Model Agents](https://mlsanity.com/paper/2502.17189/) (2025) - [Root Cause Localization for Microservice Systems in Cloud-edge Collaborative Environments](https://mlsanity.com/paper/2406.13604/) (2024) - [PRAXIS: Integrating Program Analysis with Observability for Root-Cause Analysis](https://mlsanity.com/paper/2512.22113/) (2025) ## Built on by (most cited first, 3 of 3) - [Towards a Belief-Based World Model for LLM Agents](https://mlsanity.com/paper/2609.00455/) (2026) - [A Context Engineering Framework for Improving Enterprise AI Agents based on Digital-Twin MDP](https://mlsanity.com/paper/2603.22083/) (2026) - [Verification as an Architectural Layer for LLM Agents: A V-Model Design, and a Pilot Study of Its Deterministic Core](https://mlsanity.com/paper/2609.31937/) (2026) ## Abstract (arXiv) LLM agents excel when environments are mostly static and the needed information fits in a model's context window, but they often fail in open-ended investigations where explanations must be constructed by iteratively mining evidence from massive, heterogeneous operational data. These investigations exhibit hidden dependency structure: entities interact, signals co-vary, and the importance of a fact may only become clear after other evidence is discovered. Because the context window is bounded, agents must summarize intermediate findings before their significance is known, increasing the risk of discarding key evidence. ReAct-style agents are especially brittle in this regime. Their retrieve-summarize-reason loop makes conclusions sensitive to exploration order and introduces run-to-run non-determinism, producing a reliability gap where Pass-at-k may be high but Majority-at-k remains low. Simply sampling more rollouts or generating longer reasoning traces does not reliably stabilize results, since hypotheses cannot be autonomously checked as new evidence arrives and there is no explicit mechanism for belief bookkeeping and revision. In addition, ReAct entangles semantic reasoning with controller duties such as tool orchestration and state tracking, so execution errors and plan drift degrade reasoning while consuming scarce context. We address these issues by formulating investigation as abductive reasoning over a dependency graph and proposing EoG (Explanations over Graphs), a disaggregated framework in which an LLM performs bounded local evidence mining and labeling (cause vs symptom) while a deterministic controller manages traversal, state, and belief propagation to compute a minimal explanatory frontier. On a representative ITBench diagnostics task, EoG improves both accuracy and run-to-run consistency over ReAct baselines, including a 7x average gain in Majority-at-k entity F1.