{"slug": "interpreting-black-box-large-language-models-with-sentence-level-energy", "title": "Interpreting Black-Box Large Language Models with Sentence-Level Energy Landscapes", "summary": "Researchers propose a model-agnostic, post-hoc attribution interpreter for proprietary large language models (LLMs) accessed via closed APIs, using an Energy-Based Model (EBM) as a surrogate to capture sentence-level conceptual consistency. The interpreter, once trained, quantifies the influence of prompt sentences on target outputs without further API queries, and experiments show it effectively identifies influential sentences.", "body_md": "arXiv:2608.02879v1 Announce Type: new\nAbstract: The widespread adoption of proprietary Large Language Models (LLMs) accessed strictly through closed APIs has created a critical challenge for responsible deployment: a fundamental lack of interpretability. To address this, we propose a model-agnostic, post-hoc attribution interpreter operating at the sentence level. Our approach trains an Energy-Based Model (EBM) as a surrogate to capture the LLM's internal conceptual consistency between prompts and responses. This energy landscape guides the training of a lightweight interpreter network. Uniquely, our interpreter operates as a standalone tool; once trained, it quantifies the influence of prompt sentences on a user-specified target output without requiring further API queries to the LLM. By globally training a local interpreter across diverse inputs, our framework captures broader generation patterns and mitigates instance-specific biases. Experiments demonstrate that our EBM accurately simulates the target LLM, allowing the interpreter to effectively identify the prompt sentences most influential in generating specific target outputs.", "url": "https://wpnews.pro/news/interpreting-black-box-large-language-models-with-sentence-level-energy", "canonical_source": "https://arxiv.org/abs/2608.02879", "published_at": "2026-08-05 04:00:00+00:00", "updated_at": "2026-08-05 04:09:48.945180+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research"], "entities": [], "alternates": {"html": "https://wpnews.pro/news/interpreting-black-box-large-language-models-with-sentence-level-energy", "markdown": "https://wpnews.pro/news/interpreting-black-box-large-language-models-with-sentence-level-energy.md", "text": "https://wpnews.pro/news/interpreting-black-box-large-language-models-with-sentence-level-energy.txt", "jsonld": "https://wpnews.pro/news/interpreting-black-box-large-language-models-with-sentence-level-energy.jsonld"}}