# Learning to Diagnose and Correct Moral Errors: Beyond Shallow Heuristics in Moral Alignment

arXiv: https://arxiv.org/abs/2601.03079 · HTML: https://mlsanity.com/paper/2601.03079/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 27 of its 53 references that are in the map.
- First posted on arXiv 6 Jan 2026.

## Builds on (most cited first, 27 of 27)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Jailbroken: How Does LLM Safety Training Fail?](https://mlsanity.com/paper/2307.02483/) (2023)
- [RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Models](https://mlsanity.com/paper/2009.11462/) (2020)
- [LIMA: Less Is More for Alignment](https://mlsanity.com/paper/2305.11206/) (2023)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [BBQ: A Hand-Built Bias Benchmark for Question Answering](https://mlsanity.com/paper/2110.08193/) (2021)
- [Underspecification Presents Challenges for Credibility in Modern Machine Learning](https://mlsanity.com/paper/2011.03395/) (2020)
- [Safety Alignment Should Be Made More Than Just a Few Tokens Deep](https://mlsanity.com/paper/2406.05946/) (2024)
- [Social Bias Frames: Reasoning about Social and Power Implications of Language](https://mlsanity.com/paper/1911.03891/) (2019)
- [Latent Hatred: A Benchmark for Understanding Implicit Hate Speech](https://mlsanity.com/paper/2109.05322/) (2021)
- [Evaluating the Moral Beliefs Encoded in LLMs](https://mlsanity.com/paper/2307.14324/) (2023)
- [A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicity](https://mlsanity.com/paper/2401.01967/) (2024)
- [Towards Measuring and Modeling "Culture" in LLMs: A Survey](https://mlsanity.com/paper/2403.15412/) (2024)
- [Moral Mimicry: Large Language Models Produce Moral Rationalizations Tailored to Political Identity](https://mlsanity.com/paper/2209.12106/) (2022)
- [Ethical Reasoning over Moral Alignment: A Case and Framework for In-Context Ethical Policies in LLMs](https://mlsanity.com/paper/2310.07251/) (2023)
- [Are Large Language Models Consistent over Value-laden Questions?](https://mlsanity.com/paper/2407.02996/) (2024)
- [Language Model Alignment in Multilingual Trolley Problems](https://mlsanity.com/paper/2407.02273/) (2024)
- [Normative Evaluation of Large Language Models with Everyday Moral Dilemmas](https://mlsanity.com/paper/2501.18081/) (2025)
- [MoralBERT: A Fine-Tuned Language Model for Capturing Moral Values in Social Discussions](https://mlsanity.com/paper/2403.07678/) (2024)
- [Natural Language Decompositions of Implicit Content Enable Better Text Representations](https://mlsanity.com/paper/2305.14583/) (2023)
- [Intrinsic Self-correction for Enhanced Morality: An Analysis of Internal Mechanisms and the Superficial Hypothesis](https://mlsanity.com/paper/2407.15286/) (2024)
- [MORABLES: A Benchmark for Assessing Abstract Moral Reasoning in LLMs with Fables](https://mlsanity.com/paper/2509.12371/) (2025)
- [Pragmatic Inference Chain (PIC) Improving LLMs' Reasoning of Authentic Implicit Toxic Language](https://mlsanity.com/paper/2503.01539/) (2025)
- [Diagnosing Moral Reasoning Acquisition in Language Models: Pragmatics and Generalization](https://mlsanity.com/paper/2502.16600/) (2025)
- [Position: AI Safety Must Embrace an Antifragile Perspective](https://mlsanity.com/paper/2509.13339/) (2025)
- [Discourse Heuristics For Paradoxically Moral Self-Correction](https://mlsanity.com/paper/2507.00985/) (2025)

## Abstract (arXiv)

Existing approaches to moral value alignment are primarily set out to align LLMs' generation with the distributions of morally appropriate language, which has seen good progress. However, these approaches are often brittle, heavily rely on shallow heuristics, and reduce performance in out-of-the-distribution tasks. In other words, the learning paradigm underlying existing approaches teaches LLMs what morally (in)appropriate language looks like, but not why it is morally (in)appropriate. In this paper, we address this challenge by developing pragmatic inference-driven methods to facilitate LLMs' learning of how to diagnose and correct moral errors, thereby enabling them to generate morally appropriate language. Pragmatic inference is the reasoning process of deriving (implied) meanings -- a famous concept in linguistics. Our methods vary the inference procedures by the inferential load of different moral discourses, rather than modelling their diverse and complex semantic distributions separately. Empirical results demonstrate that our approach improves moral value alignment in LLMs and generalizes effectively across tasks.