Communities
Connect sessions
AI calendar
Organizations
Join Slack
Contact Sales

Terms and Conditions

Twitter GitHub LinkedIn Bluesky Youtube

© 2026 ResearchTrend.AI, All rights reserved.

Home
Papers
2507.23221
Cited By

A Single Direction of Truth: An Observer Model's Linear Residual Probe Exposes and Steers Contextual Hallucinations

A Single Direction of Truth: An Observer Model's Linear Residual Probe Exposes and Steers Contextual Hallucinations

31 July 2025

Charles OÑeill

Mudith Jayasekara

ArXiv (abs)PDF HTML

Papers citing "A Single Direction of Truth: An Observer Model's Linear Residual Probe Exposes and Steers Contextual Hallucinations"

4 / 4 papers shown

When Bias Pretends to Be Truth: How Spurious Correlations Undermine Hallucination Detection in LLMs

When Bias Pretends to Be Truth: How Spurious Correlations Undermine Hallucination Detection in LLMs

323

0

0

10 Nov 2025

Weak Form Learning for Mean-Field Partial Differential Equations: an Application to Insect Movement

Weak Form Learning for Mean-Field Partial Differential Equations: an Application to Insect Movement

Benjamin Van Allen

120

0

0

09 Oct 2025

Painless Activation Steering: An Automated, Lightweight Approach for Post-Training Large Language Models

Painless Activation Steering: An Automated, Lightweight Approach for Post-Training Large Language Models

238

1

0

25 Sep 2025

Beyond Transcription: Mechanistic Interpretability in ASR

Beyond Transcription: Mechanistic Interpretability in ASR

Yael Segal-Feldman

96

0

0

21 Aug 2025