Saved in:
Bibliographic Details
Main Authors: Yuan, Chenchen, Zhang, Zheyu, Kasneci, Gjergji
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.03609
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918484593606656
author Yuan, Chenchen
Zhang, Zheyu
Kasneci, Gjergji
author_facet Yuan, Chenchen
Zhang, Zheyu
Kasneci, Gjergji
contents Large language models often display heterogeneous moral preferences across settings. We study inference-time steering toward a desired ethical framework while preserving general competence. We present Convergent-Divergent Routing, which traces and edits minimal branch points inside transformer blocks where ethical-framework-related pathways first converge and then diverge. Gating non-target branches at these loci blocks the downstream propagation while leaving upstream computations intact. We find that this intervention alone increases targeted ethical-framework reasoning. To achieve fine-grained control, we adapt Common Spatial Patterns to the residual stream and extract, for each branch-point layer, a pair of directions that discriminate between utilitarian and deontological frameworks. We then introduce Dual Logit Calibration, a closed-form, minimum-$\ell_2$-norm update that moves the residual within this two-dimensional subspace so the resulting directional projections align with user-specified preference weights. Experiments on real-life moral dilemmas show that our method reliably achieves preference calibration and largely preserves general capabilities, outperforming recent baselines while providing an interpretable mechanism.
format Preprint
id arxiv_https___arxiv_org_abs_2605_03609
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Where Paths Split: Localized, Calibrated Control of Moral Reasoning in Large Language Models
Yuan, Chenchen
Zhang, Zheyu
Kasneci, Gjergji
Artificial Intelligence
Machine Learning
Large language models often display heterogeneous moral preferences across settings. We study inference-time steering toward a desired ethical framework while preserving general competence. We present Convergent-Divergent Routing, which traces and edits minimal branch points inside transformer blocks where ethical-framework-related pathways first converge and then diverge. Gating non-target branches at these loci blocks the downstream propagation while leaving upstream computations intact. We find that this intervention alone increases targeted ethical-framework reasoning. To achieve fine-grained control, we adapt Common Spatial Patterns to the residual stream and extract, for each branch-point layer, a pair of directions that discriminate between utilitarian and deontological frameworks. We then introduce Dual Logit Calibration, a closed-form, minimum-$\ell_2$-norm update that moves the residual within this two-dimensional subspace so the resulting directional projections align with user-specified preference weights. Experiments on real-life moral dilemmas show that our method reliably achieves preference calibration and largely preserves general capabilities, outperforming recent baselines while providing an interpretable mechanism.
title Where Paths Split: Localized, Calibrated Control of Moral Reasoning in Large Language Models
topic Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2605.03609