MoReBench: Evaluating Procedural and Pluralistic Moral Reasoning in Language Models, More than Outcomes
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915561499262976 |
|---|---|
| author | Chiu, Yu Ying Lee, Michael S. Calcott, Rachel Handoko, Brandon de Font-Reaulx, Paul Rodriguez, Paula Zhang, Chen Bo Calvin Han, Ziwen Sehwag, Udari Madhushani Maurya, Yash Knight, Christina Q Lloyd, Harry R. Bacus, Florence Mazeika, Mantas Liu, Bing Choi, Yejin Gordon, Mitchell L Levine, Sydney |
| author_facet | Chiu, Yu Ying Lee, Michael S. Calcott, Rachel Handoko, Brandon de Font-Reaulx, Paul Rodriguez, Paula Zhang, Chen Bo Calvin Han, Ziwen Sehwag, Udari Madhushani Maurya, Yash Knight, Christina Q Lloyd, Harry R. Bacus, Florence Mazeika, Mantas Liu, Bing Choi, Yejin Gordon, Mitchell L Levine, Sydney |
| contents | As AI systems progress, we rely more on them to make decisions with us and for us. To ensure that such decisions are aligned with human values, it is imperative for us to understand not only what decisions they make but also how they come to those decisions. Reasoning language models, which provide both final responses and (partially transparent) intermediate thinking traces, present a timely opportunity to study AI procedural reasoning. Unlike math and code problems which often have objectively correct answers, moral dilemmas are an excellent testbed for process-focused evaluation because they allow for multiple defensible conclusions. To do so, we present MoReBench: 1,000 moral scenarios, each paired with a set of rubric criteria that experts consider essential to include (or avoid) when reasoning about the scenarios. MoReBench contains over 23 thousand criteria including identifying moral considerations, weighing trade-offs, and giving actionable recommendations to cover cases on AI advising humans moral decisions as well as making moral decisions autonomously. Separately, we curate MoReBench-Theory: 150 examples to test whether AI can reason under five major frameworks in normative ethics. Our results show that scaling laws and existing benchmarks on math, code, and scientific reasoning tasks fail to predict models' abilities to perform moral reasoning. Models also show partiality towards specific moral frameworks (e.g., Benthamite Act Utilitarianism and Kantian Deontology), which might be side effects of popular training paradigms. Together, these benchmarks advance process-focused reasoning evaluation towards safer and more transparent AI. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_16380 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MoReBench: Evaluating Procedural and Pluralistic Moral Reasoning in Language Models, More than Outcomes Chiu, Yu Ying Lee, Michael S. Calcott, Rachel Handoko, Brandon de Font-Reaulx, Paul Rodriguez, Paula Zhang, Chen Bo Calvin Han, Ziwen Sehwag, Udari Madhushani Maurya, Yash Knight, Christina Q Lloyd, Harry R. Bacus, Florence Mazeika, Mantas Liu, Bing Choi, Yejin Gordon, Mitchell L Levine, Sydney Computation and Language Artificial Intelligence Computers and Society Human-Computer Interaction Machine Learning As AI systems progress, we rely more on them to make decisions with us and for us. To ensure that such decisions are aligned with human values, it is imperative for us to understand not only what decisions they make but also how they come to those decisions. Reasoning language models, which provide both final responses and (partially transparent) intermediate thinking traces, present a timely opportunity to study AI procedural reasoning. Unlike math and code problems which often have objectively correct answers, moral dilemmas are an excellent testbed for process-focused evaluation because they allow for multiple defensible conclusions. To do so, we present MoReBench: 1,000 moral scenarios, each paired with a set of rubric criteria that experts consider essential to include (or avoid) when reasoning about the scenarios. MoReBench contains over 23 thousand criteria including identifying moral considerations, weighing trade-offs, and giving actionable recommendations to cover cases on AI advising humans moral decisions as well as making moral decisions autonomously. Separately, we curate MoReBench-Theory: 150 examples to test whether AI can reason under five major frameworks in normative ethics. Our results show that scaling laws and existing benchmarks on math, code, and scientific reasoning tasks fail to predict models' abilities to perform moral reasoning. Models also show partiality towards specific moral frameworks (e.g., Benthamite Act Utilitarianism and Kantian Deontology), which might be side effects of popular training paradigms. Together, these benchmarks advance process-focused reasoning evaluation towards safer and more transparent AI. |
| title | MoReBench: Evaluating Procedural and Pluralistic Moral Reasoning in Language Models, More than Outcomes |
| topic | Computation and Language Artificial Intelligence Computers and Society Human-Computer Interaction Machine Learning |
| url | https://arxiv.org/abs/2510.16380 |