Multilingual Trolley Problems for Language Models
- URL: http://arxiv.org/abs/2407.02273v1
- Date: Tue, 2 Jul 2024 14:02:53 GMT
- Title: Multilingual Trolley Problems for Language Models
- Authors: Zhijing Jin, Sydney Levine, Max Kleiman-Weiner, Giorgio Piatti, Jiarui Liu, Fernando Gonzalez Adauto, Francesco Ortu, András Strausz, Mrinmaya Sachan, Rada Mihalcea, Yejin Choi, Bernhard Schölkopf,
- Abstract summary: This study is inspired by a large-scale cross-cultural study of human moral preferences, "The Moral Machine Experiment"
We show that large language models (LLMs) are more aligned with human preferences in languages such as English, Korean, Hungarian, and Chinese, but less aligned in languages such as Hindi and Somali (in Africa)
We also characterize the explanations LLMs give for their moral choices and find that fairness is the most dominant supporting reason behind GPT-4's decisions and utilitarianism by GPT-3.
- Score: 138.0995992619116
- License: http://creativecommons.org/licenses/by-sa/4.0/
- Abstract: As large language models (LLMs) are deployed in more and more real-world situations, it is crucial to understand their decision-making when faced with moral dilemmas. Inspired by a large-scale cross-cultural study of human moral preferences, "The Moral Machine Experiment", we set up the same set of moral choices for LLMs. We translate 1K vignettes of moral dilemmas, parametrically varied across key axes, into 100+ languages, and reveal the preferences of LLMs in each of these languages. We then compare the responses of LLMs to that of human speakers of those languages, harnessing a dataset of 40 million human moral judgments. We discover that LLMs are more aligned with human preferences in languages such as English, Korean, Hungarian, and Chinese, but less aligned in languages such as Hindi and Somali (in Africa). Moreover, we characterize the explanations LLMs give for their moral choices and find that fairness is the most dominant supporting reason behind GPT-4's decisions and utilitarianism by GPT-3. We also discover "language inequality" (which we define as the model's different development levels in different languages) in a series of meta-properties of moral decision making.
Related papers
- Decoding Multilingual Moral Preferences: Unveiling LLM's Biases Through the Moral Machine Experiment [11.82100047858478]
This paper builds on the moral machine experiment (MME) to investigate the moral preferences of five large language models in a multilingual setting.
We generate 6500 scenarios of the MME and prompt the models in ten languages on which action to take.
Our analysis reveals that all LLMs inhibit different moral biases to some degree and that they not only differ from the human preferences but also across multiple languages within the models themselves.
arXiv Detail & Related papers (2024-07-21T14:48:13Z) - Ethical Reasoning and Moral Value Alignment of LLMs Depend on the Language we Prompt them in [19.675262411557235]
This paper explores how three prominent LLMs -- GPT-4, ChatGPT, and Llama2-70B-Chat -- perform ethical reasoning in different languages.
We experiment with six languages: English, Spanish, Russian, Chinese, Hindi, and Swahili.
We find that GPT-4 is the most consistent and unbiased ethical reasoner across languages, while ChatGPT and Llama2-70B-Chat show significant moral value bias when we move to languages other than English.
arXiv Detail & Related papers (2024-04-29T06:42:27Z) - Hire a Linguist!: Learning Endangered Languages with In-Context
Linguistic Descriptions [52.95579788485411]
LINGOLLM is a training-free approach to enable an LLM to process unseen languages that hardly occur in its pre-training.
We implement LINGOLLM on top of two models, GPT-4 and Mixtral, and evaluate their performance on 5 tasks across 8 endangered or low-resource languages.
Our results show that LINGOLLM elevates translation capability from GPT-4's 0 to 10.5 BLEU for 10 language directions.
arXiv Detail & Related papers (2024-02-28T03:44:01Z) - OMGEval: An Open Multilingual Generative Evaluation Benchmark for Large
Language Models [59.54423478596468]
We introduce OMGEval, the first Open-source Multilingual Generative test set that can assess the capability of LLMs in different languages.
For each language, OMGEval provides 804 open-ended questions, covering a wide range of important capabilities of LLMs.
Specifically, the current version of OMGEval includes 5 languages (i.e., Zh, Ru, Fr, Es, Ar)
arXiv Detail & Related papers (2024-02-21T04:42:41Z) - Do Moral Judgment and Reasoning Capability of LLMs Change with Language?
A Study using the Multilingual Defining Issues Test [21.108525674360898]
We extend the work of beyond English to five new languages (Chinese, Hindi, Russian, Spanish and Swahili)
Our study shows that the moral reasoning ability for all models, as indicated by the post-conventional score, is substantially inferior for Hindi and Swahili, compared to Spanish, Russian, Chinese and English.
arXiv Detail & Related papers (2024-02-03T12:52:36Z) - Are Multilingual LLMs Culturally-Diverse Reasoners? An Investigation into Multicultural Proverbs and Sayings [73.48336898620518]
Large language models (LLMs) are highly adept at question answering and reasoning tasks.
We study the ability of a wide range of state-of-the-art multilingual LLMs to reason with proverbs and sayings in a conversational context.
arXiv Detail & Related papers (2023-09-15T17:45:28Z) - The Moral Machine Experiment on Large Language Models [0.0]
This study utilized the Moral Machine framework to investigate the ethical decision-making tendencies of large language models (LLMs)
While LLMs' and humans' preferences are broadly aligned, PaLM 2 and Llama 2, especially, evidence distinct deviations.
These insights elucidate the ethical frameworks of LLMs and their potential implications for autonomous driving.
arXiv Detail & Related papers (2023-09-12T04:49:39Z) - Speaking Multiple Languages Affects the Moral Bias of Language Models [70.94372902010232]
Pre-trained multilingual language models (PMLMs) are commonly used when dealing with data from multiple languages and cross-lingual transfer.
Do the models capture moral norms from English and impose them on other languages?
Our experiments demonstrate that, indeed, PMLMs encode differing moral biases, but these do not necessarily correspond to cultural differences or commonalities in human opinions.
arXiv Detail & Related papers (2022-11-14T20:08:54Z) - Do Multilingual Language Models Capture Differing Moral Norms? [71.52261949766101]
Massively multilingual sentence representations are trained on large corpora of uncurated data.
This may cause the models to grasp cultural values including moral judgments from the high-resource languages.
The lack of data in certain languages can also lead to developing random and thus potentially harmful beliefs.
arXiv Detail & Related papers (2022-03-18T12:26:37Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.