mlsanityData to 7 Oct 2026

On the Context Sensitivity of LLM Moral Judgment

arXiv 2603.23114 ↗First posted 24 Mar 2026cs.AI cs.CL cs.CYcited by 3

Not on the map. Why?

Abstract (arXiv)

A human's moral decision depends heavily on the context. Yet research on LLM morality has largely studied fixed scenarios. We address this gap by introducing Contextual MoralChoice, a dataset of moral dilemmas with systematic contextual variations known from moral psychology to shift human judgment: consequentialist, emotional, and relational. Evaluating 22 LLMs, we find that nearly all models are context-sensitive, shifting their judgments towards rule-violating behavior. Comparing with a human survey, we find that models and humans are most triggered by different contextual variations, and that a model aligned with human judgments in the base case is not necessarily aligned in its contextual sensitivity. This raises the question of controlling contextual sensitivity, which we address with an activation steering approach that can reliably increase or decrease a model's contextual sensitivity. Code and data: https://github.com/adrian-sauter/contextual_moralchoice.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 47 of its 92 references that are in the map. First posted on arXiv 24 Mar 2026.

Builds on

47 of 92 references.

References · 47

  1. 2022Training language models to follow instructions with human feedbackcited by 15,712
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  4. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  6. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  7. 2019HellaSwag: Can a Machine Really Finish Your Sentence?cited by 4,018
  8. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 2,950
  9. 2022Locating and Editing Factual Associations in GPTcited by 2,609
  10. 2023Inference-Time Intervention: Eliciting Truthful Answers from a Language Modelcited by 1,259
  11. 2022Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 smallcited by 1,130
  12. 2023Steering Llama 2 via Contrastive Activation Additioncited by 1,009
35 more
  1. 2020Beyond Accuracy: Behavioral Testing of NLP models with CheckListcited by 979
  2. 2020MPNet: Masked and Permuted Pre-training for Language Understandingcited by 973
  3. 2024Refusal in Language Models Is Mediated by a Single Directioncited by 960
  4. 2023Enhancing Chat Language Models by Scaling High-quality Instructional Conversationscited by 825
  5. 2023Steering Language Models With Activation Engineeringcited by 824
  6. 2023Whose Opinions Do Language Models Reflect?cited by 759
  7. 2020Aligning AI With Shared Human Valuescited by 706
  8. 2023The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasetscited by 673
  9. 2022Beyond neural scaling laws: beating power law scaling via data pruningcited by 519
  10. 2024Mechanistic Interpretability for AI Safety -- A Reviewcited by 415
  11. 2018TextBugger: Generating Adversarial Text Against Real-world Applicationscited by 376
  12. 2021Measuring and Improving Consistency in Pretrained Language Modelscited by 361
  13. 2023GPT-Driver: Learning to Drive with GPTcited by 344
  14. 2023Large Language Models Sensitivity to The Order of Options in Multiple-Choice Questionscited by 249
  15. 2023In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steeringcited by 240
  16. 2023PromptRobust: Towards Evaluating the Robustness of Large Language Models on Adversarial Promptscited by 218
  17. 2023Evaluating the Moral Beliefs Encoded in LLMscited by 211
  18. 2021Large Pre-trained Language Models Contain Human-like Biases of What is Right and Wrong to Docited by 178
  19. 2023Personality Traits in Large Language Modelscited by 166
  20. 2021Can Machines Learn Morality? The Delphi Experimentcited by 140
  21. 2023Taken out of context: On measuring situational awareness in LLMscited by 131
  22. 2020Scruples: A Corpus of Community Ethical Judgments on 32,000 Real-Life Anecdotescited by 100
  23. 2022When to Make Exceptions: Exploring Language Models as Accounts of Human Moral Judgmentcited by 93
  24. 2022The Moral Integrity Corpus: A Benchmark for Ethical Dialogue Systemscited by 92
  25. 2023Ethical Reasoning over Moral Alignment: A Case and Framework for In-Context Ethical Policies in LLMscited by 64
  26. 2024Spectral Editing of Activations for Large Language Model Alignmentcited by 54
  27. 2023MoCa: Measuring Human-Language Model Alignment on Causal and Moral Judgment Taskscited by 44
  28. 2022ClarifyDelphi: Reinforced Clarification Questions with Defeasibility Rewards for Social and Moral Situationscited by 36
  29. 2024Language Models in Dialogue: Conversational Maxims for Human-AI Interactionscited by 21
  30. 2024Are Large Language Models Moral Hypocrites? A Study Based on Moral Foundationscited by 19
  31. 2024Attributions toward Artificial Agents in a modified Moral Turing Testcited by 18
  32. 2025Discerning What Matters: A Multi-Dimensional Assessment of Moral Competence in LLMscited by 10
  33. 2023Toward best research practices in AI Psychologycited by 8
  34. 2026The Fragility Of Moral Judgment In Large Language Modelscited by 8
  35. 2025More or Less Wrong: A Benchmark for Directional Bias in LLM Comparative Reasoningcited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026To Nuke or Not to Nuke: LLMs' (Missing) Ethical Reasoning and Actions in a High-Stakes Decision-Making Simulationcited by 1
  2. 2026What Does Activation Steering Control? Attribution Across Answer Encodings and Output-Sensitive Subspacescited by 1
  3. 2026Echoes of Deeds: Moral History Can Shape and Steer LLM Behavioral Choicescited by 0