Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

MORALISE: A Structured Benchmark for Moral Alignment in Visual Language Models

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A new benchmark of 2,481 real-world image-text pairs shows that vision-language models can usually tell right from wrong but often fail to name which of 13 moral norms a scenario violates.

desk verdict A useful real-image moral alignment benchmark whose headline numbers need an annotation-agreement and human-baseline check before being read as ground truth. read the letter →

arxiv 2505.14728 v1 pith:KX3AY7KJ submitted 2025-05-20 cs.CV cs.AIcs.CLcs.CYcs.MM

Xiao Lin , Zhining Liu , Ze Yang , Gaotang Li , Ruizhong Qiu , Shuke Wang , Hui Liu , Haotian Li
show 5 more authors
Sumit Keswani Vishwa Pardeshi Huijun Zhao Wei Fan Hanghang Tong
This is my paper · ORCID
classification cs.CVcs.AIcs.CLcs.CYcs.MM
keywords moralalignmentvision-languagemodelsmultimodalbenchmarknormattributionjudgmentTuriel'sdomaintheoryreal-worldimage-textdatamodalityannotation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper introduces MORALISE, a benchmark that tests whether vision-language models can tell when an image-text scenario is morally wrong and, crucially, which of 13 moral norms it violates. The authors assemble 2,481 real-world image-text pairs, each manually labeled for the violated moral topic and for whether the violation is carried by the image or the text, then evaluate 19 open-source and proprietary models. They find that models are reasonably good at binary moral judgment, averaging 88.28 accuracy for proprietary models and 83.55 for open-source models, but much worse at naming the violated norm, with average hit rates of 66.60 and 42.63 respectively. The central claim is that moral alignment in multimodal settings remains an open problem, and that fine-grained moral norm attribution—not just coarse judgment—is the binding constraint.

What carries the argument

The load-bearing mechanism is the pairing of a 13-topic moral taxonomy with modality-centric annotation. Each sample is labeled with one or more of 13 moral topics, organized into personal, interpersonal, and societal domains following Turiel's domain theory, and with a binary modality label indicating whether the moral violation is conveyed primarily through the image or through the text. This supports two evaluation tasks: moral judgment, a binary wrong/not-wrong decision, and moral norm attribution, scored as single-norm hit rate and multi-norm F1 over the 13 topics. The taxonomy gives the benchmark its fine-grained diagnostic power, while the modality label isolates where models fail.

What would settle it

Re-annotate the 2,481 image-text pairs with a second independent panel and compute per-topic and per-modality inter-annotator agreement, then run the same 19 models alongside a human baseline on the norm-attribution task. If agreement falls below conventional thresholds, such as Cohen's kappa under 0.6 on topic labels, or if human participants score no higher than the best models on the fine-grained task, then the reported moral deficiencies would be better explained by label ambiguity or task design than by model incapacity.

Watch

Extended reading notes

Core claim

On the authors' own terms, MORALISE establishes that current vision-language models exhibit a systematic gap between recognizing that something is morally wrong and identifying which moral norm is violated. Using a taxonomy of 13 moral topics rooted in Turiel's domain theory, spanning personal, interpersonal, and societal domains, the authors curated 2,481 expert-verified real-world image-text pairs with both topic and modality labels. Across 19 models, binary moral judgment accuracy averages 88.28 for proprietary and 83.55 for open-source models, but single-norm attribution hit rates drop to 66.60 and 42.63 respectively, and multi-label F1 scores are lower still. Even the strongest evaluated model, GPT-4o, reaches only 42.32 F1 on the 'respect' topic, and the paper reads this as evidence that fine-grained moral reasoning is a distinct and largely unsolved capability.

Load-bearing premise

The benchmark's ground-truth labels, which say which scenarios are morally wrong, which of 13 topics they violate, and whether the violation is image-centric or text-centric, are taken as correct on the strength of a majority vote among graduate student annotators, with no reported inter-annotator agreement, external validation of the taxonomy, or human performance baseline.

Editorial extensions

If this is right

  • Fine-grained moral norm attribution is substantially harder for vision-language models than binary moral judgment: the average gap is roughly 22 accuracy points for proprietary models and 41 for open-source models.
  • Text-centric moral violations are consistently easier than image-centric ones across all three subtasks, indicating that current VLMs lean on language rather than visual content when reasoning about morality.
  • Scaling model size helps moral performance up to about 10 billion parameters and then plateaus, so larger models alone will not close the gap without targeted moral-alignment training.
  • Norms that are common in social discourse, such as harm, justice, and integrity, are handled relatively well, while abstract norms such as liberty, respect, and reciprocity remain persistent weak points, especially in multi-label attribution.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If label noise is a real driver, then reporting human performance on the same 2,481 pairs would calibrate the benchmark: a human baseline near or below GPT-4o's 42 F1 on 'respect' would suggest the task is ambiguous rather than the model deficient.
  • The text/image gap points to a concrete testable extension: pipelines that first caption the image and then judge morality from the caption should recover most of the text-centric advantage, directly testing whether image understanding or moral reasoning is the bottleneck.
  • A fixed 13-topic taxonomy may encode culturally specific moral assumptions, so a comparative study with annotators from different cultural backgrounds could reveal whether the norm-attribution failures partly reflect a cultural mismatch rather than a general moral deficit.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper introduces MORALISE, a benchmark for evaluating moral alignment in vision-language models (VLMs). It proposes a taxonomy of 13 moral topics organized under Turiel's Domain Theory (personal, interpersonal, societal), curates 2,481 real-world image-text pairs with topic and modality (text-centric vs. image-centric) annotations, and defines two tasks: moral judgment (binary) and moral norm attribution (single- and multi-label). The authors evaluate 19 open and proprietary VLMs across three subtasks (S1 binary accuracy, S2 single-norm hit rate, S3 multi-norm F1) and report large performance drops from binary judgment to fine-grained attribution, along with analyses of model scale, family, modality sensitivity, and prediction correlation. The benchmark is publicly released on Hugging Face.

Significance. If the benchmark labels are reliable, MORALISE would be a valuable resource: it is one of the first VLM moral-alignment benchmarks to use real-world images rather than AI-generated ones, covers a relatively broad taxonomy, provides modality-centric annotations that enable isolating visual vs. textual moral cues, and evaluates a substantial set of models. The finding that proprietary and open-source models retain high binary judgment accuracy but drop sharply on norm attribution (e.g., proprietary average 88.28 vs. 66.60, open-source 83.55 vs. 42.63) is a potentially important and actionable observation for alignment research. The paper also commendably provides the full dataset and detailed prompts for reproducibility. However, the current manuscript does not establish the reliability of its ground-truth labels, so these claims are not yet fully supported.

major comments (4)
  1. [§3.2, Appendix E]
  2. [§3.1, Tables 3–4]
  3. [§4.2, Tables 2–4]
  4. [§3.2, §5]
minor comments (6)
  1. [References]
  2. [Tables 5–7]
  3. [§3.2, Appendix A]
  4. [Appendix B.1, Prompt τS3]
  5. [§4.4, Figure 6]
  6. [§1, §4.1]

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the benchmark evaluates models against external human annotations, with no fitted parameters or self-citation chain.

full rationale

MORALISE is a dataset-and-evaluation benchmark, not a derivation. The gold labels are produced by a majority-vote annotation protocol among graduate students (Section 3.2) and are independent of the evaluated VLMs. The central result, that models score lower on norm attribution than on binary moral judgment, is an empirical measurement against those fixed labels, not a quantity fitted from model outputs. The taxonomy is grounded in Turiel's Domain Theory and prior cited work, none of which is authored by the present authors, and no uniqueness theorem or self-citation is invoked to rule out alternatives. The absence of inter-annotator agreement statistics or a human baseline is a legitimate validity concern, but it concerns whether the labels are correct, not whether the paper's analysis reduces to its own inputs. Appendix E's acknowledgment that the curation pipeline is labor-intensive likewise describes scalability, not circularity. No equation or definition in the paper defines the target quantity in terms of the fitted inputs, and no model is trained on the benchmark before evaluation. The paper therefore exhibits no self-definitional, fitted-input, self-citation, or renaming circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

There are no fitted numerical parameters and no invented physical entities. The benchmark rests on domain assumptions about the moral taxonomy and the quality of human annotation, plus the ad hoc assumption that the 13 topics are separable enough for the attribution tasks.

assumptions (3)
  • domain assumption Turiel's Domain Theory provides a valid basis for organizing moral values into personal, interpersonal, and societal domains.
    Section 3.1 builds the 13-topic taxonomy entirely on this theory. If the theory or its application to multimodal scenarios is invalid, the benchmark categories may not correspond to real moral distinctions.
  • domain assumption Graduate student annotators' majority-vote labels are accurate ground truth for moral violation, topic, and modality.
    Section 3.2 relies on annotator majority vote without reporting inter-annotator agreement. Incorrect or inconsistent labels would directly bias model scores and the paper's conclusions.
  • ad hoc to paper The 13 moral topics are sufficiently distinct that single-norm attribution is a meaningful task.
    The definitions overlap substantially (Justice, Fairness, Discrimination), and the paper itself uses multi-label annotations. Penalizing a model that selects a near-synonym topic may understate its moral reasoning ability.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MORALISE: A Structured Benchmark for Moral Alignment in Visual Language Models." pith.science (2026). https://pith.science/paper/KX3AY7KJ

@misc{pith2026250514728,
  author       = {Pith},
  title        = {Pith review of: MORALISE: A Structured Benchmark for Moral Alignment in Visual Language Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KX3AY7KJ}},
  note         = {Machine review of arXiv:2505.14728}
}
read the original abstract

Warning: This paper contains examples of harmful language and images. Reader discretion is advised. Recently, vision-language models have demonstrated increasing influence in morally sensitive domains such as autonomous driving and medical analysis, owing to their powerful multimodal reasoning capabilities. As these models are deployed in high-stakes real-world applications, it is of paramount importance to ensure that their outputs align with human moral values and remain within moral boundaries. However, existing work on moral alignment either focuses solely on textual modalities or relies heavily on AI-generated images, leading to distributional biases and reduced realism. To overcome these limitations, we introduce MORALISE, a comprehensive benchmark for evaluating the moral alignment of vision-language models (VLMs) using diverse, expert-verified real-world data. We begin by proposing a comprehensive taxonomy of 13 moral topics grounded in Turiel's Domain Theory, spanning the personal, interpersonal, and societal moral domains encountered in everyday life. Built on this framework, we manually curate 2,481 high-quality image-text pairs, each annotated with two fine-grained labels: (1) topic annotation, identifying the violated moral topic(s), and (2) modality annotation, indicating whether the violation arises from the image or the text. For evaluation, we encompass two tasks, \textit{moral judgment} and \textit{moral norm attribution}, to assess models' awareness of moral violations and their reasoning ability on morally salient content. Extensive experiments on 19 popular open- and closed-source VLMs show that MORALISE poses a significant challenge, revealing persistent moral limitations in current state-of-the-art models. The full benchmark is publicly available at https://huggingface.co/datasets/Ze1025/MORALISE.

Figures

Figures reproduced from arXiv: 2505.14728 by the authors.

Figure 1
Figure 1. Overview of the proposed MORALISE benchmark. Best viewed in color. and compassion by responding to others’ needs and suffering; (4) Harm: Avoiding actions that cause physical or emotional injury to others; (5) Fairness: Distributing resources or opportunities impartially, without favoritism or bias; (6) Reciprocity: Returning favors and cooperation fairly when others offer help; (7) Loyalty: Staying faithful to one’… view at source ↗
Figure 2
Figure 2. Representative examples for all 13 moral topics and two modality-centric violations. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Impact of model size on moral alignment. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Topic-level model average performance comparison. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Moral sensitivity to modality-centric violations. Across all subtasks, we plot distributions [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Prediction correlation across model architectures. RQ7: Do models from the same family exhibit simi￾lar behavior? Finally, we conducted a correlation anal￾ysis on model outputs to examine whether moral con￾cepts are consistently represented across different models. The…
Figure 7
Figure 7. Figure 7: Radar plots of dataset statistics. The left and middle plots illustrate the distribution of [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: Heatmap analysis on the similarity of model moral predictions. [PITH_FULL_IMAGE:figures/full_fig_p017_8.png]
Figure 9
Figure 9. Figure 9: Detailed model comparison for moral judgement. Models’ performance has been rescaled [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]
Figure 10
Figure 10. Figure 10: Detailed model comparison for single-norm attribution. Models’ performance has been [PITH_FULL_IMAGE:figures/full_fig_p020_10.png]
Figure 11
Figure 11. Figure 11: Detailed model comparison for multi-norm attribution. Models’ performance has been [PITH_FULL_IMAGE:figures/full_fig_p021_11.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. NormAct: Benchmarking Embodied Agents' Proactive Compliance with Unspoken Social Norms

    cs.AI 2026-06 unverdicted novelty 6.5 of 10

    NormAct shows MLLMs reach explicit goals in 67.3% of cases but comply with hidden norms in only 26.4%, with NormPerceptor raising task success from 24.2% to 46.7%.

Reference graph

Works this paper leans on

55 extracted references · 22 canonical work pages · cited by 1 Pith paper

  1. [1]

    https://openai.com/index/ o3-o4-mini-system-card/

    OpenAI o3 and o4-mini System Card — openai.com. https://openai.com/index/ o3-o4-mini-system-card/. [Accessed 10-05-2025]

    work page 2025

  2. [2]

    Phi-3 technical report: A highly capable language model locally on your phone, 2024

    Marah Abdin, Jyoti Aneja, Hany Awadalla, Ahmed Awadallah, Ammar Ahmad Awan, Nguyen Bach, Amit Bahree, Arash Bakhtiari, Jianmin Bao, Harkirat Behl, Alon Benhaim, Misha Bilenko, Johan Bjorck, Sébastien Bubeck, Martin Cai, Qin Cai, Vishrav Chaudhary, Dong Chen, Dongdong Chen, et al. Phi-3 technical report: A highly capable language model locally on your phone, 2024

    work page 2024

  3. [3]

    Moral foundations of large language models

    Marwa Abdulhai, Gregory Serapio-García, Clément Crepy, Daria Valter, John Canny, and Natasha Jaques. Moral foundations of large language models. In Yaser Al-Onaizan, Mohit Bansal, and Yun-Nung Chen, editors,Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024, Miami, FL, USA, November 12-16, 2024, pages 17737–...

    work page 2024

  4. [4]

    Gpt-4 technical report.arXiv preprint arXiv:2303.08774, 2023

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report.arXiv preprint arXiv:2303.08774, 2023

    arXiv 2023

  5. [5]

    Flamingo: a visual language model for few-shot learning.Advances in neural information processing systems, 35:23716–23736, 2022

    Jean-Baptiste Alayrac, Jeff Donahue, Pauline Luc, Antoine Miech, Iain Barr, Yana Hasson, Karel Lenc, Arthur Mensch, Katherine Millican, Malcolm Reynolds, et al. Flamingo: a visual language model for few-shot learning.Advances in neural information processing systems, 35:23716–23736, 2022

    2022

  6. [6]

    Qwen2.5-vl technical report, 2025

    Shuai Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Sibo Song, Kai Dang, Peng Wang, Shijie Wang, Jun Tang, Humen Zhong, Yuanzhi Zhu, Mingkun Yang, Zhaohai Li, Jianqiang Wan, Pengfei Wang, Wei Ding, Zheren Fu, Yiheng Xu, Jiabo Ye, Xi Zhang, Tianbao Xie, Zesen Cheng, Hang Zhang, Zhibo Yang, Haiyang Xu, and Junyang Lin. Qwen2.5-vl technical report, 2025

    2025

  7. [7]

    Physbench: Benchmarking and enhancing vision-language models for physical world understanding.arXiv preprint arXiv:2501.16411, 2025

    Wei Chow, Jiageng Mao, Boyi Li, Daniel Seita, Vitor Guizilini, and Yue Wang. Physbench: Benchmarking and enhancing vision-language models for physical world understanding.arXiv preprint arXiv:2501.16411, 2025

    arXiv 2025

  8. [8]

    Hwang, Vered Shwartz, Maarten Sap, and Yejin Choi

    Maxwell Forbes, Jena D. Hwang, Vered Shwartz, Maarten Sap, and Yejin Choi. Social chemistry 101: Learning to reason about social and moral norms. In Bonnie Webber, Trevor Cohn, Yulan He, and Yang Liu, editors,Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 653–670, Online, November 2020. Association fo...

    work page 2020

Show all 55 references
  1. [9]

    Bias and fairness in large language models: A survey.Computational Linguistics, 50(3):1097–1179, 2024

    Isabel O Gallegos, Ryan A Rossi, Joe Barrow, Md Mehrab Tanjim, Sungchul Kim, Franck Dernoncourt, Tong Yu, Ruiyi Zhang, and Nesreen K Ahmed. Bias and fairness in large language models: A survey.Computational Linguistics, 50(3):1097–1179, 2024

    2024

  2. [10]

    Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning.arXiv preprint arXiv:2501.12948, 2025

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning.arXiv preprint arXiv:2501.12948, 2025

    2025 arXiv

  3. [11]

    Vision-language models for medical report generation and visual question answering: a review.Frontiers Artif

    Iryna Hartsock and Ghulam Rasool. Vision-language models for medical report generation and visual question answering: a review.Frontiers Artif. Intell., 7, 2024

    2024

  4. [12]

    Aligning ai with shared human values.arXiv preprint arXiv:2008.02275, 2020

    Dan Hendrycks, Collin Burns, Steven Basart, Andrew Critch, Jerry Li, Dawn Song, and Jacob Steinhardt. Aligning ai with shared human values.arXiv preprint arXiv:2008.02275, 2020

    2008 arXiv

  5. [13]

    VIV A: A benchmark for vision-grounded decision- making with human values

    Zhe Hu, Yixiao Ren, Jing Li, and Yu Yin. VIV A: A benchmark for vision-grounded decision- making with human values. In Yaser Al-Onaizan, Mohit Bansal, and Yun-Nung Chen, editors, Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024...

    2024

  6. [14]

    Moralbench: Moral evaluation of llms.arXiv preprint arXiv:2406.04428, 2024

    Jianchao Ji, Yutong Chen, Mingyu Jin, Wujiang Xu, Wenyue Hua, and Yongfeng Zhang. Moralbench: Moral evaluation of llms.arXiv preprint arXiv:2406.04428, 2024

    2024 arXiv

  7. [15]

    Sadhu, Zhuochun Li, Xizhi Wu, Shyam Visweswaran, and Yanshan Wang

    Yuelyu Ji, Wenhe Ma, Sonish Sivarajkumar, Hang Zhang, Eugene M. Sadhu, Zhuochun Li, Xizhi Wu, Shyam Visweswaran, and Yanshan Wang. Mitigating the risk of health inequity exacerbated by large language models.npj Digital Medicine, 8(1):246, 2025

    2025

  8. [16]

    Hwang, Chandra Bhagavatula, Ronan Le Bras, Jenny T

    Liwei Jiang, Jena D. Hwang, Chandra Bhagavatula, Ronan Le Bras, Jenny T. Liang, Sydney Levine, Jesse Dodge, Keisuke Sakaguchi, Maxwell Forbes, Jack Hessel, Jonathan Borchardt, Taylor Sorensen, Saadia Gabriel, Yulia Tsvetkov, Oren Etzioni, Maarten Sap, Regina Rini, and Yejin Ch...

    2025

  9. [17]

    Gemma 3 technical report.CoRR, abs/2503.19786, 2025

    Aishwarya Kamath, Johan Ferret, Shreya Pathak, Nino Vieillard, Ramona Merhej, Sarah Perrin, Tatiana Matejovicova, Alexandre Ramé, Morgane Rivière, Louis Rouillard, Thomas Mesnard, Geoffrey Cideron, Jean-Bastien Grill, Sabela Ramos, Edouard Yvinec, Michelle Casbon, Etienne Pot,...

    2025 arXiv

  10. [18]

    who’s in charge?

    Marta Laupa. “who’s in charge?” preschool children’s concepts of authority.Early Childhood Research Quarterly, 9(1):1–17, 1994

    1994

  11. [19]

    VHELM: A holistic evaluation of vision language models

    Tony Lee, Haoqin Tu, Chi Heem Wong, Wenhao Zheng, Yiyang Zhou, Yifan Mai, Jos- selin Somerville Roberts, Michihiro Yasunaga, Huaxiu Yao, Cihang Xie, and Percy Liang. VHELM: A holistic evaluation of vision language models. In Amir Globersons, Lester Mackey, Danielle Belgrave, A...

    2024

  12. [20]

    Blip: Bootstrapping language- image pre-training for unified vision-language understanding and generation

    Junnan Li, Dongxu Li, Caiming Xiong, and Steven Hoi. Blip: Bootstrapping language- image pre-training for unified vision-language understanding and generation. InInternational conference on machine learning, pages 12888–12900. PMLR, 2022

    2022

  13. [21]

    Benchmark evalua- tions, applications, and challenges of large vision language models: A survey.arXiv preprint arXiv:2501.02189, 1, 2025

    Zongxia Li, Xiyang Wu, Hongyang Du, Huy Nghiem, and Guangyao Shi. Benchmark evalua- tions, applications, and challenges of large vision language models: A survey.arXiv preprint arXiv:2501.02189, 1, 2025

    2025 arXiv

  14. [22]

    Llava-next: Improved reasoning, ocr, and world knowledge, January 2024

    Haotian Liu, Chunyuan Li, Yuheng Li, Bo Li, Yuanhan Zhang, Sheng Shen, and Yong Jae Lee. Llava-next: Improved reasoning, ocr, and world knowledge, January 2024

    2024

  15. [23]

    Visual instruction tuning, 2023

    Haotian Liu, Chunyuan Li, Qingyang Wu, and Yong Jae Lee. Visual instruction tuning, 2023

    2023

  16. [24]

    Learn to explain: Multimodal reasoning via thought chains for science question answering

    Pan Lu, Swaroop Mishra, Tanglin Xia, Liang Qiu, Kai-Wei Chang, Song-Chun Zhu, Oyvind Tafjord, Peter Clark, and Ashwin Kalyan. Learn to explain: Multimodal reasoning via thought chains for science question answering. In Sanmi Koyejo, S. Mohamed, A. Agarwal, Danielle Belgrave, K...

    2022

  17. [25]

    Stereoset: Measuring stereotypical bias in pretrained language models.arXiv preprint arXiv:2004.09456, 2020

    Moin Nadeem, Anna Bethke, and Siva Reddy. Stereoset: Measuring stereotypical bias in pretrained language models.arXiv preprint arXiv:2004.09456, 2020

    2004 arXiv

  18. [26]

    Aylward, Marc Edgar, Michael Zephyr, Pavlo Molchanov, Baris Turkbey, Holger Roth, and Daguang Xu

    Vishwesh Nath, Wenqi Li, Dong Yang, Andriy Myronenko, Mingxin Zheng, Yao Lu, Zhijian Liu, Hongxu Yin, Yee Man Law, Yucheng Tang, Pengfei Guo, Can Zhao, Ziyue Xu, Yufan He, Greg Heinrich, Stephen R. Aylward, Marc Edgar, Michael Zephyr, Pavlo Molchanov, Baris Turkbey, Holger Rot...

    2024 arXiv

  19. [27]

    Autonomy and the personal: Negotiation and social reciprocity in adult-child social exchanges.New Directions for Child and Adolescent Development, 1996(73):7–24, 1996

    Larry P Nucci, Melanie Killen, and Judith G Smetana. Autonomy and the personal: Negotiation and social reciprocity in adult-child social exchanges.New Directions for Child and Adolescent Development, 1996(73):7–24, 1996. 11

    1996

  20. [28]

    Gpt-4o system card, 2024

    OpenAI. Gpt-4o system card, 2024

    2024

  21. [29]

    VLP: vision language planning for autonomous driving

    Chenbin Pan, Burhaneddin Yaman, Tommaso Nesti, Abhirup Mallik, Alessandro Gabriele Allievi, Senem Velipasalar, and Liu Ren. VLP: vision language planning for autonomous driving. InIEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024, Seattle, WA, USA, June...

    2024

  22. [30]

    Learning transferable visual models from natural language supervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervision. InInternational conference on machine learning, pag...

    2021

  23. [31]

    Biasdora: Exploring hidden biased associations in vision-language models.arXiv preprint arXiv:2407.02066, 2024

    Chahat Raj, Anjishnu Mukherjee, Aylin Caliskan, Antonios Anastasopoulos, and Ziwei Zhu. Biasdora: Exploring hidden biased associations in vision-language models.arXiv preprint arXiv:2407.02066, 2024

    2024 arXiv

  24. [32]

    Alignment in multimodal interaction: An integrative framework.Cognitive science, 44(11):e12911, 2020

    Marlou Rasenberg, Asli Özyürek, and Mark Dingemanse. Alignment in multimodal interaction: An integrative framework.Cognitive science, 44(11):e12911, 2020

    2020

  25. [33]

    Children’s recognition of fairness and others’ welfare in a resource allocation task: Age related changes.Developmental psychology, 52(8):1307, 2016

    Michael T Rizzo, Laura Elenbaas, Shelby Cooley, and Melanie Killen. Children’s recognition of fairness and others’ welfare in a resource allocation task: Age related changes.Developmental psychology, 52(8):1307, 2016

    2016

  26. [34]

    Object hallucination in image captioning.arXiv preprint arXiv:1809.02156, 2018

    Anna Rohrbach, Lisa Anne Hendricks, Kaylee Burns, Trevor Darrell, and Kate Saenko. Object hallucination in image captioning.arXiv preprint arXiv:1809.02156, 2018

    2018 arXiv

  27. [35]

    Evaluating the moral beliefs encoded in llms.Advances in Neural Information Processing Systems, 36:51778–51809, 2023

    Nino Scherrer, Claudia Shi, Amir Feder, and David Blei. Evaluating the moral beliefs encoded in llms.Advances in Neural Information Processing Systems, 36:51778–51809, 2023

    2023

  28. [37]

    Assessment of multimodal large language models in alignment with human values.arXiv preprint arXiv:2403.17830, 2024

    Zhelun Shi, Zhipin Wang, Hongxing Fan, Zaibin Zhang, Lijun Li, Yongting Zhang, Zhenfei Yin, Lu Sheng, Yu Qiao, and Jing Shao. Assessment of multimodal large language models in alignment with human values.arXiv preprint arXiv:2403.17830, 2024

    2024 arXiv

  29. [38]

    Enhanc- ing the learning experience: Using vision-language models to generate questions for educational videos, 2025

    Markos Stamatakis, Joshua Berger, Christian Wartena, Ralph Ewerth, and Anett Hoppe. Enhanc- ing the learning experience: Using vision-language models to generate questions for educational videos, 2025

    2025

  30. [39]

    Ryutaro Tanno, David G. T. Barrett, Andrew Sellergren, Sumedh Ghaisas, Sumanth Dathathri, Abigail See, Johannes Welbl, Karan Singhal, Shekoofeh Azizi, Tao Tu, Mike Schaekermann, Rhys May, Roy Lee, SiWai Man, Zahra Ahmed, S. Sara Mahdavi, Danielle Belgrave, Vivek Natarajan, Shr...

    2023 arXiv

  31. [40]

    Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context.arXiv preprint arXiv:2403.05530, 2024

    Gemini Team, Petko Georgiev, Ving Ian Lei, Ryan Burnell, Libin Bai, Anmol Gulati, Garrett Tanzer, Damien Vincent, Zhufeng Pan, Shibo Wang, et al. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context.arXiv preprint arXiv:2403.05530, 2024

    2024 arXiv

  32. [41]

    Qwen3, April 2025

    Qwen Team. Qwen3, April 2025

    2025

  33. [42]

    Drivevlm: The convergence of autonomous driving and large vision-language models.arXiv preprint arXiv:2402.12289, 2024

    Xiaoyu Tian, Junru Gu, Bailin Li, Yicheng Liu, Yang Wang, Zhiyong Zhao, Kun Zhan, Peng Jia, Xianpeng Lang, and Hang Zhao. Drivevlm: The convergence of autonomous driving and large vision-language models.arXiv preprint arXiv:2402.12289, 2024

    2024 arXiv

  34. [43]

    Mothers’ and teachers’ home and school rules: Young children’s conceptions of authority in context.Merrill- Palmer Quarterly (1982-), pages 168–187, 2000

    Marie S Tisak, Dushka Crane-Ross, John Tisak, and Amanda M Maynard. Mothers’ and teachers’ home and school rules: Young children’s conceptions of authority in context.Merrill- Palmer Quarterly (1982-), pages 168–187, 2000

    1982

  35. [44]

    Cambridge University Press, 1983

    Elliot Turiel.The development of social knowledge: Morality and convention. Cambridge University Press, 1983. 12

    1983

  36. [45]

    Qwen2-vl: Enhancing vision- language model’s perception of the world at any resolution, 2024

    Peng Wang, Shuai Bai, Sinan Tan, Shijie Wang, Zhihao Fan, Jinze Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Yang Fan, Kai Dang, Mengfei Du, Xuancheng Ren, Rui Men, Dayiheng Liu, Chang Zhou, Jingren Zhou, and Junyang Lin. Qwen2-vl: Enhancing vision- language model’s p...

    2024

  37. [46]

    M3oralbench: A multi- modal moral benchmark for lvlms.CoRR, abs/2412.20718, 2024

    Bei Yan, Jie Zhang, Zhiyuan Chen, Shiguang Shan, and Xilin Chen. M3oralbench: A multi- modal moral benchmark for lvlms.CoRR, abs/2412.20718, 2024

    2024 arXiv

  38. [47]

    Safebench: A safety evaluation framework for multimodal large language models.arXiv preprint arXiv:2410.18927, 2024

    Zonghao Ying, Aishan Liu, Siyuan Liang, Lei Huang, Jinyang Guo, Wenbo Zhou, Xianglong Liu, and Dacheng Tao. Safebench: A safety evaluation framework for multimodal large language models.arXiv preprint arXiv:2410.18927, 2024

    2024 arXiv

  39. [48]

    Chatglm: A family of large language models from GLM-130B to GLM-4 all tools.CoRR, abs/2406.12793, 2024

    Aohan Zeng, Bin Xu, Bowen Wang, Chenhui Zhang, Da Yin, Diego Rojas, Guanyu Feng, Hanlin Zhao, Hanyu Lai, Hao Yu, Hongning Wang, Jiadai Sun, Jiajie Zhang, Jiale Cheng, Jiayi Gui, Jie Tang, Jing Zhang, Juanzi Li, Lei Zhao, Lindong Wu, Lucen Zhong, Mingdao Liu, Minlie Huang, Peng...

    2024 arXiv

  40. [49]

    Vision-language models for vision tasks: A survey.IEEE Trans

    Jingyi Zhang, Jiaxing Huang, Sheng Jin, and Shijian Lu. Vision-language models for vision tasks: A survey.IEEE Trans. Pattern Anal. Mach. Intell., 46(8):5625–5644, 2024

    2024

  41. [50]

    Spa-vl: A comprehensive safety preference alignment dataset for vision language model.arXiv preprint arXiv:2406.12030, 2024

    Yongting Zhang, Lu Chen, Guodong Zheng, Yifeng Gao, Rui Zheng, Jinlan Fu, Zhenfei Yin, Senjie Jin, Yu Qiao, Xuanjing Huang, et al. Spa-vl: A comprehensive safety preference alignment dataset for vision language model.arXiv preprint arXiv:2406.12030, 2024

    2024 arXiv

  42. [51]

    Towards foundation-model- based multiagent system to accelerate ai for social impact.AAMAS, 2025

    Yunfan Zhao, Niclas Boehmer, Aparna Taneja, and Milind Tambe. Towards foundation-model- based multiagent system to accelerate ai for social impact.AAMAS, 2025

    2025

  43. [52]

    Vlmbench: A com- positional benchmark for vision-and-language manipulation.Advances in Neural Information Processing Systems, 35:665–678, 2022

    Kaizhi Zheng, Xiaotong Chen, Odest Chadwicke Jenkins, and Xin Wang. Vlmbench: A com- positional benchmark for vision-and-language manipulation.Advances in Neural Information Processing Systems, 35:665–678, 2022

    2022

  44. [53]

    Vlstereoset: A study of stereotypical bias in pre- trained vision-language models

    Kankan Zhou, Yibin LAI, and Jing Jiang. Vlstereoset: A study of stereotypical bias in pre- trained vision-language models. Association for Computational Linguistics, 2022

    2022

  45. [54]

    Vision language models in autonomous driving: A survey and outlook.IEEE Transactions on Intelligent Vehicles, 2024

    Xingcheng Zhou, Mingyu Liu, Ekim Yurtsever, Bare Luka Zagar, Walter Zimmer, Hu Cao, and Alois C Knoll. Vision language models in autonomous driving: A survey and outlook.IEEE Transactions on Intelligent Vehicles, 2024

    2024

  46. [55]

    Internvl3: Exploring advanced training and test-time recipes for open-source multimodal models, 2025

    Jinguo Zhu, Weiyun Wang, Zhe Chen, Zhaoyang Liu, Shenglong Ye, Lixin Gu, et al. Internvl3: Exploring advanced training and test-time recipes for open-source multimodal models, 2025

    2025

  47. [56]

    The moral integrity corpus: A benchmark for ethical dialogue systems.arXiv preprint arXiv:2204.03021, 2022

    Caleb Ziems, Jane A Yu, Yi-Chia Wang, Alon Halevy, and Diyi Yang. The moral integrity corpus: A benchmark for ethical dialogue systems.arXiv preprint arXiv:2204.03021, 2022. 13 A Dataset Statistics discrimination integrity loyalty reciprocityfairness responsibility sanctity li...

    2022 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.