Agentic Interpretation: Lattice-Structured Evidence for LLM-Based Program Analysis

• URL: http://arxiv.org/abs/2605.12694v1
• Date: Tue, 12 May 2026 19:46:24 GMT
• Title: Agentic Interpretation: Lattice-Structured Evidence for LLM-Based Program Analysis
• Abstract summary: We propose agentic interpretation, a framework that brings the discipline of lattice-based static analysis to program reasoning.<n>A worklist algorithm governs how claims and their judgments evolve during the analysis.
• Score: 2.1833756646187763
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models can consult information that fixed static analyzers cannot, such as documentation, current security advisories, version-specific metadata, and informal API contracts. This makes LLMs a compelling option for program analyses that depend on information beyond the source program, or that are otherwise not amenable to conventional static analyzers. However, directly asking an LLM for a one-shot whole-program analysis is brittle because it compresses many evidence-dependent judgments into a single opaque answer, rather than exposing which conclusions are supported or disputed and using intermediate findings to guide later, more focused searches. In this paper, we propose agentic interpretation, a framework that brings the discipline of lattice-based static analysis to LLM-driven program reasoning. At a high level, agentic interpretation decomposes a high-level analysis goal into localized claims, and tracks the LLM's judgment about each claim in a finite-height lattice. A worklist algorithm governs how claims and their judgments evolve during the analysis. We introduce a formal model of agentic interpretation, explore the design space it opens, and illustrate the approach with a worked example analyzing code that depends on opaque third-party components.

Related papers

• Framing Migration News with LLMs: Structured CoT as a Support for Human Interpretation [7.86841540422233]
  We introduce a Structured Chain-of-Thought (SCoT) prompting approach using Llama3-8B.<n>SCoT allows users to audit model outputs and examine alternative interpretations in a task that is inherently subjective.<n>We conduct a human-centered evaluation in which annotators assess the coherence and influence of "the model's reasoning"
  arXiv  Detail & Related papers  (2026-06-02T15:15:39Z)
• LongDA: Benchmarking LLM Agents for Long-Document Data Analysis [55.32211515932351]
  LongDA targets real-world settings in which navigating long documentation and complex data is the primary bottleneck.<n>LongTA is a tool-augmented agent framework that enables document access, retrieval, and code execution.<n>Our experiments reveal substantial performance gaps even among state-of-the-art models.
  arXiv  Detail & Related papers  (2026-01-05T23:23:16Z)
• Computational frame analysis revisited: On LLMs for studying news coverage [1.4528491369411618]
  Generative LLMs like GPT and Claude are increasingly being used as content analytical tools.<n>We systematically evaluate them against their computational predecessors.<n>We conclude by endorsing a methodologically pluralistic approach and put forth a roadmap for computational frame analysis for researchers going forward.
  arXiv  Detail & Related papers  (2025-11-21T19:52:46Z)
• IDA-Bench: Evaluating LLMs on Interactive Guided Data Analysis [60.32962597618861]
  IDA-Bench is a novel benchmark evaluating large language models in multi-round interactive scenarios.<n>Agent performance is judged by comparing its final numerical output to the human-derived baseline.<n>Even state-of-the-art coding agents (like Claude-3.7-thinking) succeed on 50% of the tasks, highlighting limitations not evident in single-turn tests.
  arXiv  Detail & Related papers  (2025-05-23T09:37:52Z)
• Flowco: Rethinking Data Analysis in the Age of LLMs [2.1874189959020427]
  Large language models (LLMs) are now capable of generating such code for simple, routine analyses.<n>LLMs promise to democratize data science by enabling those with limited programming expertise to conduct data analyses.<n>Analysts in many real-world settings must often exercise fine-grained control over specific analysis steps.<n>This paper introduces Flowco, a new mixed-initiative system to address these challenges.
  arXiv  Detail & Related papers  (2025-04-18T19:01:27Z)
• How do Large Language Models Understand Relevance? A Mechanistic Interpretability Perspective [64.00022624183781]
  Large language models (LLMs) can assess relevance and support information retrieval (IR) tasks.<n>We investigate how different LLM modules contribute to relevance judgment through the lens of mechanistic interpretability.
  arXiv  Detail & Related papers  (2025-04-10T16:14:55Z)
• Understanding Formal Reasoning Failures in LLMs as Abstract Interpreters [5.468061853910803]
  We introduce two novel prompting strategies that aim to elicit such reasoning from large language models (LLMs)<n>We evaluate these strategies across several state-of-the-art LLMs on 22 programs from the SV-COMP benchmark suite widely used in software verification.
  arXiv  Detail & Related papers  (2025-03-16T23:05:52Z)
• Investigating the Shortcomings of LLMs in Step-by-Step Legal Reasoning [34.427730009102966]
  We develop an automated evaluation framework to identify reasoning errors and evaluate the performance of LLMs.<n>Our work will also serve as an evaluation framework that can be used in detailed error analysis of reasoning chains for logic-intensive complex tasks.
  arXiv  Detail & Related papers  (2025-02-08T19:49:32Z)
• LLMSA: A Compositional Neuro-Symbolic Approach to Compilation-free and Customizable Static Analysis [13.993290878789779]
  We propose a compositional neuro-symbolic approach for compilation-free, customizable static analysis with reduced hallucinations.<n>It attains 66.27% precision and 78.57% recall in taint vulnerability detection, surpassing an industrial approach in F1 score by 0.20.
  arXiv  Detail & Related papers  (2024-12-18T23:14:59Z)
• Interactive Topic Models with Optimal Transport [75.26555710661908]
  We present EdTM, as an approach for label name supervised topic modeling. EdTM models topic modeling as an assignment problem while leveraging LM/LLM based document-topic affinities.
  arXiv  Detail & Related papers  (2024-06-28T13:57:27Z)
• The Emergence of Large Language Models in Static Analysis: A First Look through Micro-Benchmarks [3.848607479075651]
  We investigate the role that current Large Language Models (LLMs) can play in improving callgraph analysis and type inference for Python programs. Our study reveals that LLMs show promising results in type inference, demonstrating higher accuracy than traditional methods, yet they exhibit limitations in callgraph analysis.
  arXiv  Detail & Related papers  (2024-02-27T16:53:53Z)
• LLM Inference Unveiled: Survey and Roofline Model Insights [62.92811060490876]
  Large Language Model (LLM) inference is rapidly evolving, presenting a unique blend of opportunities and challenges. Our survey stands out from traditional literature reviews by not only summarizing the current state of research but also by introducing a framework based on roofline model. This framework identifies the bottlenecks when deploying LLMs on hardware devices and provides a clear understanding of practical problems.
  arXiv  Detail & Related papers  (2024-02-26T07:33:05Z)
• Sentiment Analysis through LLM Negotiations [58.67939611291001]
  A standard paradigm for sentiment analysis is to rely on a singular LLM and makes the decision in a single round. This paper introduces a multi-LLM negotiation framework for sentiment analysis.
  arXiv  Detail & Related papers  (2023-11-03T12:35:29Z)

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.