# Normative Robustness as a Frontier for Non-Verifiable Reasoning in LLMs

arXiv: https://arxiv.org/abs/2606.12731 · HTML: https://mlsanity.com/paper/2606.12731/

## 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 26 of its 85 references that are in the map.
- First posted on arXiv 10 Jun 2026.

## Placed in

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

## Builds on (most cited first, 26 of 26)

- [Self-Consistency Improves Chain of Thought Reasoning in Language Models](https://mlsanity.com/paper/2203.11171/) (2022)
- [Jailbroken: How Does LLM Safety Training Fail?](https://mlsanity.com/paper/2307.02483/) (2023)
- [A General Language Assistant as a Laboratory for Alignment](https://mlsanity.com/paper/2112.00861/) (2021)
- [Aligning AI With Shared Human Values](https://mlsanity.com/paper/2008.02275/) (2020)
- ["Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models](https://mlsanity.com/paper/2308.03825/) (2023)
- [The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"](https://mlsanity.com/paper/2309.12288/) (2023)
- [ARC-AGI-2: A New Challenge for Frontier AI Reasoning Systems](https://mlsanity.com/paper/2505.11831/) (2025)
- [Can Machines Learn Morality? The Delphi Experiment](https://mlsanity.com/paper/2110.07574/) (2021)
- [Large Language Models Often Know When They Are Being Evaluated](https://mlsanity.com/paper/2505.23836/) (2025)
- [SycEval: Evaluating LLM Sycophancy](https://mlsanity.com/paper/2502.08177/) (2025)
- [Towards Robust Mathematical Reasoning](https://mlsanity.com/paper/2511.01846/) (2025)
- [Alice in Wonderland: Simple Tasks Showing Complete Reasoning Breakdown in State-Of-the-Art Large Language Models](https://mlsanity.com/paper/2406.02061/) (2024)
- [Are Large Language Models Consistent over Value-laden Questions?](https://mlsanity.com/paper/2407.02996/) (2024)
- [Using Counterfactual Tasks to Evaluate the Generality of Analogical Reasoning in Large Language Models](https://mlsanity.com/paper/2402.08955/) (2024)
- [How do Large Language Models Navigate Conflicts between Honesty and Helpfulness?](https://mlsanity.com/paper/2402.07282/) (2024)
- [MoReBench: Evaluating Procedural and Pluralistic Moral Reasoning in Language Models, More than Outcomes](https://mlsanity.com/paper/2510.16380/) (2025)
- [Characterizing Delusional Spirals through Human-LLM Chat Logs](https://mlsanity.com/paper/2603.16567/) (2026)
- [Sycophantic Chatbots Cause Delusional Spiraling, Even in Ideal Bayesians](https://mlsanity.com/paper/2602.19141/) (2026)
- [Let's Do a Thought Experiment: Using Counterfactuals to Improve Moral Reasoning](https://mlsanity.com/paper/2306.14308/) (2023)
- [People readily follow personal advice from AI but it does not improve their well-being](https://mlsanity.com/paper/2511.15352/) (2025)
- [Evaluating and Improving Robustness in Large Language Models: A Survey and Future Directions](https://mlsanity.com/paper/2506.11111/) (2025)
- [Moral Susceptibility and Robustness under Persona Role-Play in Large Language Models](https://mlsanity.com/paper/2511.08565/) (2025)
- [Moral Sycophancy in Vision Language Models](https://mlsanity.com/paper/2602.08311/) (2026)
- [Are Language Models Sensitive to Morally Irrelevant Distractors?](https://mlsanity.com/paper/2602.09416/) (2026)
- [The Topological Trouble With Transformers](https://mlsanity.com/paper/2604.17121/) (2026)
- [Think Before You Lie: How Reasoning Leads to Honesty](https://mlsanity.com/paper/2603.09957/) (2026)

## Abstract (arXiv)

As LLMs increasingly serve in advisory and deliberative roles, users rely on them for non-verifiable reasoning in domains lacking objective ground truths. However, traditional evaluations of LLM reasoning focus almost exclusively on fact-based domains, such as mathematics and science, leaving uncertainty over whether and to what degree models can handle ambiguous, subjective, or value-laden problems over time. To address this concern, we propose moral reasoning as a paradigmatic subdomain of non-verifiable reasoning. We define moral robustness as a model's capacity to exhibit sound moral reasoning across time and contexts, and we introduce a scalable, adversarial, multi-turn evaluation framework to empirically measure this capability. We simulate 48,000 user-agent moral deliberations across four frontier LLMs, varying premise relevance, premise order, conversation duration, and the user's stated moral view. We find that models successfully ignore morally-irrelevant distractors, but shift their reasoning by up to 6.5%, on average, towards the user's stated preferred moral view, and varying their reasoning depending on factors such as order (altering moral judgments by order in 13-22% of the cases) and duration (altering moral judgments between single-turn and multi-turn in 10-24% of the cases). Our analysis indicates that models tailor not just their final verdicts but their underlying justifications to align with a user's moral viewpoint - a failure mode we characterize as moral deliberative sycophancy.