EUVIMEDEUVIMED entdecken
Crossref

KnowThyself: An Agentic Assistant for LLM Interpretability

Suraj Prasai, Mengnan Du, Ying Zhang, Fan Yang

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (48) · 41661-41663

DOI: 10.1609/aaai.v40i48.42373 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Suraj Prasai und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We develop KnowThyself, an agentic assistant that advances large language model (LLM) interpretability. Existing tools provide useful insights but remain fragmented and code-intensive. KnowThyself consolidates these capabilities into a chat-based interface, where users can upload models, pose natural language questions, and obtain interactive visualizations with guided explanations. At its core, an orchestrator LLM first reformulates user queries, an agent router further directs them to specialized modules, and the outputs are finally contextualized into coherent explanations. This design lowers technical barriers and provides an extensible platform for LLM inspection. By embedding the whole process into a conversational workflow, KnowThyself offers a robust foundation for accessible LLM interpretability.

Zitieren

APA

Prasai, S., Du, M., Zhang, Y., & Yang, F. (2026). KnowThyself: An Agentic Assistant for LLM Interpretability. Proceedings of the AAAI Conference on Artificial Intelligence, 40(48), 41661-41663. https://doi.org/10.1609/aaai.v40i48.42373

Vancouver

Prasai S, Du M, Zhang Y, Yang F. KnowThyself: An Agentic Assistant for LLM Interpretability. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(48):41661-41663. doi:10.1609/aaai.v40i48.42373

BibTeX
@article{Prasai2026,
  author = {Prasai, Suraj and Du, Mengnan and Zhang, Ying and Yang, Fan},
  title = {KnowThyself: An Agentic Assistant for LLM Interpretability},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {48},
  pages = {41661--41663},
  doi = {10.1609/aaai.v40i48.42373}
}
Kostenlos starten →