Published November 26, 2025 | Version v1
Poster Open

(Poster) The Trilemma of Truth in Large Language Models @ Mech Interp Workshop (NeurIPS 2025)

Authors/Creators

  • 1. EDMO icon Northeastern University
  • 2. ROR icon Santa Fe Institute

Description

This poster is presented at the Mechanistic Interpretability Workshop as part of the NeurIPS 2025.

Files

NeurIPS_ToT_Poster_v2.pdf

Files (743.4 kB)

Name Size Download all
md5:96407be0cdeecc35e940be8eb6eb8669
743.4 kB Preview Download

Additional details

Is new version of
Poster: 10.5281/zenodo.16907693 (DOI)
Poster: 10.5281/zenodo.16039538 (DOI)
Is supplement to
Working paper: https://openreview.net/forum?id=z7dLG2ycRf (URL)
Preprint: 10.48550/arXiv.2506.23921 (DOI)
Software: 10.5281/zenodo.16039538 (DOI)
Dataset: 10.57967/hf/5900 (DOI)
Preprint: 10.2139/ssrn.5331504 (DOI)