# Where Paths Split: Localized, Calibrated Control of Moral Reasoning in Large Language Models

arXiv: https://arxiv.org/abs/2605.03609 · HTML: https://mlsanity.com/paper/2605.03609/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 15 of its 52 references that are in the map.
- First posted on arXiv 5 May 2026.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 15 of 15)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Steering Language Models With Activation Engineering](https://mlsanity.com/paper/2308.10248/) (2023)
- [Yi: Open Foundation Models by 01.AI](https://mlsanity.com/paper/2403.04652/) (2024)
- [Persona Vectors: Monitoring and Controlling Character Traits in Language Models](https://mlsanity.com/paper/2507.21509/) (2025)
- [A Roadmap to Pluralistic Alignment](https://mlsanity.com/paper/2402.05070/) (2024)
- [DailyDilemmas: Revealing Value Preferences of LLMs with Quandaries of Daily Life](https://mlsanity.com/paper/2410.02683/) (2024)
- [AI PERSONA: Towards Life-long Personalization of LLMs](https://mlsanity.com/paper/2412.13103/) (2024)
- [Improved Representation Steering for Language Models](https://mlsanity.com/paper/2505.20809/) (2025)
- [Structured Moral Reasoning in Language Models: A Value-Grounded Evaluation Framework](https://mlsanity.com/paper/2506.14948/) (2025)
- [Addressing Moral Uncertainty using Large Language Models for Ethical Decision-Making](https://mlsanity.com/paper/2503.05724/) (2025)
- [MidPO: Dual Preference Optimization for Safety and Helpfulness in Large Language Models via a Mixture of Experts Framework](https://mlsanity.com/paper/2506.02460/) (2025)
- ["Pull or Not to Pull?'': Investigating Moral Biases in Leading Large Language Models Across Ethical Dilemmas](https://mlsanity.com/paper/2508.07284/) (2025)
- [Is ETHICS about ethics? Evaluating the ETHICS benchmark](https://mlsanity.com/paper/2410.13009/) (2024)

## Abstract (arXiv)

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-ℓ2-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.