How Post-Training Reshapes LLMs: A Mechanistic View on Knowledge, Truthfulness, Refusal, and Confidence | NSF Public Access Repository

skip to main content

An official website of the United States government Here's how you know

Official websites use .gov

A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS

A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Citation Details

How Post-Training Reshapes LLMs: A Mechanistic View on Knowledge, Truthfulness, Refusal, and Confidence

Award ID(s):: 2238714

PAR ID:: 10637455

Author(s) / Creator(s):: Du, Hongzhe; Li, Weikai; Cai, Min; Saraipour, Karim; Zhang, Zimin; Lakkaraju, Himabindu; Sun, Yizhou; Zhang, Shichang

Publisher / Repository:: Conference on Language Modeling (COLM)

Date Published:: 2025-10-07

Format(s):: Medium: X

Sponsoring Org:: National Science Foundation

Free Publicly Accessible Full Text
Accepted Manuscript
Conference Paper:
The DOI is not currently available.