Language-Specific Layer Matters: Efficient Multilingual Enhancement for Large Vision-Language Models

Bei Li; Jingbo Zhu; Lei Huang; Tong Xiao; Xiaocheng Feng; Yilin Wang; Yongyu Mu; Yuchun Fan

REVIEW 2 major objections 4 minor 52 references

Language-Specific Layer Matters: Efficient Multilingual Enhancement for Large Vision-Language Models

T0 review · 2 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The multilingual understanding of large vision-language models is carried by language-specific neuron activations in shallow layers, so fine-tuning only those layers on question-translation pairs improves multilingual ability while updating

desk verdict The abstract sells a precise layer-selection method but doesn't show the control that proves precision matters. read the letter →

arxiv 2508.18381 v1 pith:M3KU4I2M submitted 2025-08-25 cs.CL

Yuchun Fan , Yilin Wang , Yongyu Mu , Lei Huang , Bei Li , Xiaocheng Feng , Tong Xiao , Jingbo Zhu This is my paper

classification cs.CL

keywords largevision-languagemodelsmultilingualenhancementneuronactivationanalysisshallowlayersparameter-efficientfine-tuninglayerselectionquestion-translationpairslow-resourcelanguages

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

Large vision-language models respond unevenly across languages, and this paper tries to show that the source is identifiable: multilingual ability tracks language-specific neuron activations concentrated in shallow layers, the layers closest to the input. From that correlation the authors build PLAST, a training recipe that first measures those activations to choose the relevant layers and then fine-tunes only those layers using question-translation pairs. On two multilingual benchmarks, MM-Bench and MMMB, the recipe improves multilingual performance while changing only about 14% of parameters, and it also transfers to low-resource languages and complex visual reasoning. A sympathetic reader would care because it suggests a cheap, targeted way to add languages to large vision-language models without full fine-tuning or large new multilingual datasets.

What carries the argument

PLAST (Precise LAnguage-Specific layers fine-Tuning) is the central mechanism. It has two steps: monitor neuron activations across layers to identify which shallow layers respond specifically to a language, then fine-tune exactly those layers with question-translation pairs. The activation signal is what converts a correlation into a layer-selection rule; the translation-pair fine-tuning is the intervention that supposedly strengthens multilingual alignment. The argument stands or falls on whether the monitored activations pick out the layers that actually carry multilingual ability.

What would settle it

Run an ablation that silences the identified language-specific shallow-layer neurons on held-out multilingual benchmarks: if scores stay roughly the same, the claimed layer-selection mechanism is not the cause. A complementary check compares activation changes from full-model fine-tuning; if unselected layers change as much as selected ones, the selection signal is not doing the causal work.

Watch

Extended reading notes

Core claim

The authors' central claim is that the multilingual understanding of large vision-language models is not smeared across all parameters but is tied to language-specific neuron activations in shallow layers. They introduce a criterion: layers whose activations track a language are the layers responsible for that language's visual understanding. Using this criterion, the PLAST recipe selects and fine-tunes only those layers on question-translation pairs, aligning the visual and the multilingual language information. The paper reports improved multilingual performance on MM-Bench and MMMB with approximately 14% of parameters tuned, with the same recipe working for low-resource languages and comp

Load-bearing premise

The method assumes that the measured neuron activity is actually what makes the model multilingual, not just correlated with multilingual ability; if it is only a side effect, fine-tuning exactly those layers will not reliably improve performance.

Editorial extensions

If this is right

Multilingual enhancement no longer requires full-model fine-tuning: tuning roughly 14% of parameters is the reported cost for the gains seen on MM-Bench and MMMB.
Layer selection by activation monitoring offers a principled alternative to heuristic choices about which parts of a vision-language model to adapt for a new language.
The recipe's reported transfer to low-resource languages and complex visual reasoning implies the same shallow-layer mechanism matters beyond high-resource multilingual benchmarks.
Because the fine-tuning pairs are questions with translations, the method ties language alignment to visual question understanding rather than to text-only translation.

Reading between the lines

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

If the shallow-layer activations are the mechanism rather than a side effect, the same activation-based layer map could be reused across vision-language backbones without re-measuring, an extension the paper does not test.
A direct causal test the paper leaves open: silence or ablate the identified language-specific neurons and check whether multilingual benchmarks drop; if they do not, the correlation is not the mechanism.
The translation-pair recipe could be steered by activation strength: concentrate new translated data on languages whose shallow-layer signatures are faintest, targeting exactly where the model is most monolingual.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Short version: the abstract is a promissory note, not a result. The one genuinely interesting idea is to use language-specific neuron activations to locate the layers that matter for multilingual competence, then fine-tune only those. That's a cheap and testable recipe, and 14% of parameters is a real efficiency number. But the abstract doesn't show the control every critic will ask for: fine-tuning the same 14% of layers chosen at random, with the same question-translation data. Without that control, the 'precise' part of PLAST is not established. The correlation between shallow-layer activations and multilingual ability could be real but epiphenomenal; any busy layer would light up. If random shallow layers work as well, the gains come from the data budget, not the selection, and the paper's contribution shrinks to 'question-translation tuning of a small parameter subset.'

What's here that deserves credit: the observation that multilingual ability seems to live in shallow layers is worth reporting; the method is simple and reproducible in principle; the authors name MM-Bench and MMMB, which are standard, and they mention low-resource and complex visual reasoning generalization. If the full paper has actual numbers, baselines, and the random-layer control, it could be a solid incremental PEFT paper.

The soft spots are obvious. No numbers, no baselines, no error bars in the abstract, so I can't check anything. The causal step from correlation to selection is exactly where the stress-test points. Also, the abstract doesn't say whether the same translation pairs are used for layer identification and fine-tuning; if they are, that's a circularity worth checking.

Bottom line: this is a paper for people working on multilingual LVLMs and parameter-efficient tuning. It deserves a serious referee if the full text provides the missing controls. As an editor, I'd send it to review rather than desk reject, but I'd expect the reviewers to demand the random-layer baseline and the identification/fine-tuning data split. I wouldn't cite it from the abstract.

Referee Report

2 major / 4 minor

Summary. The paper proposes PLAST, a parameter-efficient fine-tuning recipe for large vision-language models (LVLMs) that targets language-specific layers. The authors report a correlation between multilingual understanding and language-specific neuron activations in shallow layers, and use this correlation to select layers for fine-tuning with question-translation pairs. The abstract claims improved multilingual performance on MM-Bench and MMMB with only 14% of parameters tuned, and generalizability to low-resource and complex visual reasoning tasks.

Significance. If the central claim is validated, the work offers a practical approach to improving multilingual capability in LVLMs at low computational cost, which would be useful for deployment in multilingual settings. However, the abstract alone provides no quantitative evidence and does not establish that the proposed layer-selection mechanism is causally necessary for the observed gains. The significance is therefore conditional on the full experimental details.

major comments (2)

[Abstract] The abstract moves from 'a salient correlation' between language-specific neuron activations and multilingual ability to the assertion that these layers are 'precisely fine-tuned' to achieve multilingual alignment. This is a causal/necessity claim unsupported by correlation evidence. The manuscript must include a control condition that fine-tunes the same parameter budget (14%) in randomly selected shallow layers, or layers selected by an alternative criterion, using the same question-translation pairs. Without such a baseline, the improvement could be driven by the data and parameter budget rather than by the 'precise' layer selection mechanism.
[Abstract] No quantitative results are reported. The abstract names MM-Bench and MMMB but gives no accuracy numbers, error bars, or comparisons to a standard baseline (e.g., full fine-tuning, LoRA, or random-layer tuning). The claim that PLAST 'effectively improves' multilingual capabilities is thus unverifiable from the abstract. At minimum, report the absolute scores and the improvement over a strong baseline in the abstract, or refer to a specific table.

minor comments (4)

[Abstract] The term 'shallow layers' is not defined. Specify which layers (e.g., early transformer blocks, visual encoder layers) are considered shallow, as this is central to the method.
[Abstract] Clarify what 'language-specific neuron activations' means operationally. How are 'language-specific' neurons identified, and what threshold is used? The selection threshold is a free parameter that could affect the 14% parameter claim.
[Abstract] The phrase 'question-translation pairs' is vague. Are these machine-translated or human-annotated? Which languages are covered? How many pairs? This information is important for reproducibility.
[Abstract] The final sentence, 'facilitating the language-specific visual information engagement in shallow layers,' is vague and interpretive. State what evidence would show this engagement and how it is measured.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity identified from the available text.

full rationale

The available manuscript excerpt, which consists of the abstract, presents PLAST as a two-stage recipe: first identify language-specific layers by monitoring neuron activations, then fine-tune only those layers with question-translation pairs, and finally evaluate on MM-Bench and MMMB. There is no equation, fitted parameter, or self-citation quoted that reduces a claimed prediction to its own input. The correlation between activations and multilingual ability is used as a selection signal, not as the evaluation metric, and the evaluation benchmarks are external to the selection procedure. Even if the causal sufficiency of the layer-selection step were questionable—for example, if randomly selected shallow layers with the same parameter budget might perform similarly—that would be a correctness or experimental-design concern, not circularity. The abstract does not define the target result in terms of the inputs, nor does it import a uniqueness result from the authors' prior work, nor does it rename a known pattern as a new derivation. Based on the evidence provided, no load-bearing circular step can be exhibited.

Assumptions & free parameters 2 free parameters · 2 assumptions · 1 invented entities

The central claim rests on the existence and utility of language-specific neurons as a causal handle on multilingual ability. The free parameters are the unspecified thresholds and training hyperparameters that determine which 14% of parameters are tuned.

free parameters (2)

Language-specific layer identification threshold
The method requires a threshold on neuron-activation statistics to classify layers as language-specific; the abstract does not give the threshold, so the selection of which layers are tuned is a free choice.
Fine-tuning hyperparameters
Standard training choices such as learning rate, batch size, and number of steps are not specified in the abstract.

assumptions (2)

domain assumption Language-specific neuron activations in shallow layers are a reliable indicator of the model's multilingual understanding ability.
The entire method rests on this correlation being meaningful and stable; asserted in the abstract's first sentence.
domain assumption Fine-tuning only the identified layers with question-translation pairs improves multilingual alignment without breaking other abilities.
The paper assumes that cross-lingual alignment from translation pairs will transfer to benchmark performance; stated as 'these layers are then precisely fine-tuned'.

invented entities (1)

Language-specific neurons
purpose: Serve as a signal to locate layers relevant to a particular language
The paper posits that some neurons specialize in language-specific processing; this is a construct used to select layers, not an independently verified anatomical or mechanistic unit.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Language-Specific Layer Matters: Efficient Multilingual Enhancement for Large Vision-Language Models." pith.science (2026). https://pith.science/paper/M3KU4I2M

@misc{pith2026250818381,
  author       = {Pith},
  title        = {Pith review of: Language-Specific Layer Matters: Efficient Multilingual Enhancement for Large Vision-Language Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M3KU4I2M}},
  note         = {Machine review of arXiv:2508.18381}
}

read the original abstract

Large vision-language models (LVLMs) have demonstrated exceptional capabilities in understanding visual information with human languages but also exhibit an imbalance in multilingual capabilities. In this work, we delve into the multilingual working pattern of LVLMs and identify a salient correlation between the multilingual understanding ability of LVLMs and language-specific neuron activations in shallow layers. Building on this insight, we introduce PLAST, a training recipe that achieves efficient multilingual enhancement for LVLMs by Precise LAnguage-Specific layers fine-Tuning. PLAST first identifies layers involved in multilingual understanding by monitoring language-specific neuron activations. These layers are then precisely fine-tuned with question-translation pairs to achieve multilingual alignment. Our empirical results on MM-Bench and MMMB demonstrate that PLAST effectively improves the multilingual capabilities of LVLMs and achieves significant efficiency with only 14% of the parameters tuned. Further analysis reveals that PLAST can be generalized to low-resource and complex visual reasoning tasks, facilitating the language-specific visual information engagement in shallow layers.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

52 extracted references · 22 canonical work pages

[1]

Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, et al. 2023. Qwen technical report. arXiv preprint arXiv:2309.16609

arXiv 2023
[2]

Jinze Bai, Shuai Bai, Shusheng Yang, Shijie Wang, Sinan Tan, Peng Wang, Junyang Lin, Chang Zhou, and Jingren Zhou. 2024. https://openreview.net/forum?id=qrGjFJVl3m Qwen- VL : A versatile vision-language model for understanding, localization, text reading, and beyond

work page 2024
[3]

Samyadeep Basu, Martin Grayson, Cecily Morrison, Besmira Nushi, Soheil Feizi, and Daniela Massiceti. 2024. https://openreview.net/forum?id=s63dtq0mwA Understanding information storage and transfer in multi-modal large language models . In The Thirty-eighth Annual Conference on Neural Information Processing Systems

2024
[4]

Thapliyal, Idan Szpektor, Julien Amelot, Xi Chen, and Radu Soricut

Soravit Changpinyo, Linting Xue, Michal Yarom, Ashish V. Thapliyal, Idan Szpektor, Julien Amelot, Xi Chen, and Radu Soricut. 2023. MaXM : Towards multilingual visual question answering. In Findings of the Association for Computational Linguistics: EMNLP

work page 2023
[5]

Lin Chen, Jinsong Li, Xiaoyi Dong, Pan Zhang, Conghui He, Jiaqi Wang, Feng Zhao, and Dahua Lin. 2024. Sharegpt4v: Improving large multi-modal models with better captions. In European Conference on Computer Vision, pages 370--387. Springer

2024
[6]

Xi Chen, Xiao Wang, Soravit Changpinyo, A. J. Piergiovanni, Piotr Padlewski, Daniel Salz, Sebastian Goodman, Adam Grycner, Basil Mustafa, Lucas Beyer, Alexander Kolesnikov, Joan Puigcerver, Nan Ding, Keran Rong, Hassan Akbari, Gaurav Mishra, Linting Xue, Ashish V. Thapliyal, James Bradbury, and Weicheng Kuo. 2023 a . https://openreview.net/forum?id=mWVoBz...

work page 2023
[7]

Xi Chen, Xiao Wang, Soravit Changpinyo, A. J. Piergiovanni, Piotr Padlewski, Daniel Salz, Sebastian Goodman, Adam Grycner, Basil Mustafa, Lucas Beyer, Alexander Kolesnikov, Joan Puigcerver, Nan Ding, Keran Rong, Hassan Akbari, Gaurav Mishra, Linting Xue, Ashish V. Thapliyal, James Bradbury, and Weicheng Kuo. 2023 b . https://openreview.net/forum?id=mWVoBz...

work page 2023
[8]

Gonzalez, Ion Stoica, and Eric P

Wei-Lin Chiang, Zhuohan Li, Zi Lin, Ying Sheng, Zhanghao Wu, Hao Zhang, Lianmin Zheng, Siyuan Zhuang, Yonghao Zhuang, Joseph E. Gonzalez, Ion Stoica, and Eric P. Xing. 2023. https://lmsys.org/blog/2023-03-30-vicuna/ Vicuna: An open-source chatbot impressing gpt-4 with 90\

2023

Show all 52 references

[9]

OpenCompass Contributors. 2023. Opencompass: A universal evaluation platform for foundation models. https://github.com/open-compass/opencompass

2023
[10]

Damai Dai, Li Dong, Yaru Hao, Zhifang Sui, Baobao Chang, and Furu Wei. 2022. https://doi.org/10.18653/V1/2022.ACL-LONG.581 Knowledge neurons in pretrained transformers . In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long ...

2022 doi
[11]

Yuchun Fan, Yongyu Mu, Yilin Wang, Lei Huang, Junhao Ruan, Bei Li, Tong Xiao, Shujian Huang, Xiaocheng Feng, and Jingbo Zhu. 2025. https://api.semanticscholar.org/CorpusID:275342393 Slam: Towards efficient multilingual reasoning via selective language alignment . In Internatio...

2025
[12]

Chaoyou Fu, Peixian Chen, Yunhang Shen, Yulei Qin, Mengdan Zhang, Xu Lin, Zhenyu Qiu, Wei Lin, Jinrui Yang, Xiawu Zheng, Ke Li, Xing Sun, and Rongrong Ji. 2023. https://doi.org/10.48550/ARXIV.2306.13394 MME: A comprehensive evaluation benchmark for multimodal large language mo...

2023 doi arXiv
[13]

Gregor Geigle, Florian Schneider, Carolin Holtermann, Chris Biemann, Radu Timofte, Anne Lauscher, and Goran Glavavs. 2025. https://api.semanticscholar.org/CorpusID:275405715 Centurio: On drivers of multilingual ability of large vision-language model . In Proceedings of XYZ Conference

2025
[14]

Qingyan Guo, Rui Wang, Junliang Guo, Bei Li, Kaitao Song, Xu Tan, Guoqing Liu, Jiang Bian, and Yujiu Yang. 2023. Connecting large language models with evolutionary algorithms yields powerful prompt optimizers. arXiv preprint arXiv:2309.08532

2023 arXiv
[15]

Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen - Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen

Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen - Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. https://openreview.net/forum?id=nZeVKeeFYf9 Lora: Low-rank adaptation of large language models . In The Tenth International Conference on Learning Representat...

2022
[16]

Lei Huang, Xiaocheng Feng, Weitao Ma, Yuchun Fan, Xiachong Feng, Yuxuan Gu, Yangfan Ye, Liang Zhao, Weihong Zhong, Baoxin Wang, Dayong Wu, Guoping Hu, Lingpeng Kong, Tong Xiao, Ting Liu, and Bing Qin. 2025 a . https://aclanthology.org/2025.acl-long.1199/ Alleviating hallucinat...

2025
[17]

Lei Huang, Xiaocheng Feng, Weitao Ma, Yuchun Fan, Xiachong Feng, Yangfan Ye, Weihong Zhong, Yuxuan Gu, Baoxin Wang, Dayong Wu, Guoping Hu, and Bing Qin. 2025 b . https://aclanthology.org/2025.acl-long.826/ Improving contextual faithfulness of large language models via retrieva...

2025
[18]

Lei Huang, Xiaocheng Feng, Weitao Ma, Yuxuan Gu, Weihong Zhong, Xiachong Feng, Weijiang Yu, Weihua Peng, Duyu Tang, Dandan Tu, and Bing Qin. 2024 a . https://doi.org/10.18653/V1/2024.FINDINGS-ACL.838 Learning fine-grained grounded citations for attributed large language models...

2024 doi
[19]

Lei Huang, Xiaocheng Feng, Weitao Ma, Liang Zhao, Yuchun Fan, Weihong Zhong, Dongliang Xu, Qing Yang, Hongtao Liu, and Bing Qin. 2024 b . https://doi.org/10.18653/V1/2024.EMNLP-MAIN.223 Advancing large language model attribution through self-improving . In Proceedings of the 2...

2024 doi
[20]

Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, and Ting Liu. 2025 c . https://doi.org/10.1145/3703155 A survey on hallucination in large language models: Principles, taxonomy, challenges, an...

2025 doi
[21]

Gabriel Ilharco, Mitchell Wortsman, Ross Wightman, Cade Gordon, Nicholas Carlini, Rohan Taori, Achal Dave, Vaishaal Shankar, Hongseok Namkoong, John Miller, Hannaneh Hajishirzi, Ali Farhadi, and Ludwig Schmidt. 2021. https://doi.org/10.5281/zenodo.5143773 Openclip (0.1)

2021 doi
[22]

Bohao Li, Yuying Ge, Yixiao Ge, Guangzhi Wang, Rui Wang, Ruimao Zhang, and Ying Shan. 2024. https://api.semanticscholar.org/CorpusID:271963485 Seed-bench: Benchmarking multimodal large language models . 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)...

2024
[23]

Lei Li, Yuwei Yin, Shicheng Li, Liang Chen, Peiyi Wang, Shuhuai Ren, Mukai Li, Yazheng Yang, Jingjing Xu, Xu Sun, Lingpeng Kong, and Qi Liu. 2023. M ^3 it: A large-scale dataset towards multi-modal multilingual instruction tuning. arXiv preprint arXiv:2306.04387

2023 arXiv
[25]

Haotian Liu, Chunyuan Li, Yuheng Li, and Yong Jae Lee. 2024 a . https://doi.org/10.1109/CVPR52733.2024.02484 Improved baselines with visual instruction tuning . In IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2024, Seattle, WA, USA, June 16-22, 2024 , p...

2024
[26]

Haotian Liu, Chunyuan Li, Yuheng Li, Bo Li, Yuanhan Zhang, Sheng Shen, and Yong Jae Lee. 2024 b . https://llava-vl.github.io/blog/2024-01-30-llava-next/ Llava-next: Improved reasoning, ocr, and world knowledge

2024
[27]

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

2023
[28]

Yuan Liu, Haodong Duan, Yuanhan Zhang, Bo Li, Songyang Zhang, Wangbo Zhao, Yike Yuan, Jiaqi Wang, Conghui He, Ziwei Liu, Kai Chen, and Dahua Lin. 2024 c . https://doi.org/10.1007/978-3-031-72658-3\_13 Mmbench: Is your multi-modal model an all-around player? In Computer Vision ...

2024 doi
[29]

Pan Lu, Swaroop Mishra, Tanglin Xia, Liang Qiu, Kai - Wei Chang, Song - Chun Zhu, Oyvind Tafjord, Peter Clark, and Ashwin Kalyan. 2022. http://papers.nips.cc/paper\_files/paper/2022/hash/11332b6b6cf4485b84afadb1352d3a9a-Abstract-Conference.html Learn to explain: Multimodal rea...

2022
[30]

Shaker, Salman H

Muhammad Maaz, Hanoona Abdul Rasheed, Abdelrahman M. Shaker, Salman H. Khan, Hisham Cholakkal, Rao Muhammad Anwer, Tim Baldwin, Michael Felsberg, and Fahad Shahbaz Khan. 2024. https://doi.org/10.48550/ARXIV.2402.14818 PALO: A polyglot large multimodal model for 5b people . CoR...

2024 doi arXiv
[31]

Yongyu Mu, Abudurexiti Reheman, Zhiquan Cao, Yuchun Fan, Bei Li, Yinqiao Li, Tong Xiao, Chunliang Zhang, and Jingbo Zhu. 2023. https://doi.org/10.18653/V1/2023.FINDINGS-ACL.653 Augmenting large language model translators via translation memories . In Findings of the Associatio...

2023 doi
[32]

Nostalgebraist. 2020. https://www.lesswrong.com/posts/AcKRB8wDpdaN6v6ru/interpreting-gpt-the-logit-lens Interpreting gpt: The logit lens . LessWrong

2020
[33]

OpenAI. 2023. https://doi.org/10.48550/ARXIV.2303.08774 GPT-4 technical report . CoRR, abs/2303.08774

2023 doi arXiv
[34]

Libo Qin, Qiguang Chen, Fuxuan Wei, Shijue Huang, and Wanxiang Che. 2023. https://doi.org/10.18653/V1/2023.EMNLP-MAIN.163 Cross-lingual prompting: Improving zero-shot chain-of-thought reasoning across languages . In Proceedings of the 2023 Conference on Empirical Methods in Na...

2023 doi
[35]

Xiaoye Qu, Mingyang Song, Wei Wei, Jianfeng Dong, and Yu Cheng. 2024. https://doi.org/10.48550/ARXIV.2408.00550 Mitigating multilingual hallucination in large vision-language models . CoRR, abs/2408.00550

2024 doi arXiv
[36]

Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever. 2021. http://proceedings.mlr.press/v139/radford21a.html Learning transferable visual models...

2021
[37]

Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, and Yuxiong He. 2020. https://doi.org/10.1109/SC41405.2020.00024 Zero: memory optimizations toward training trillion parameter models . In Proceedings of the International Conference for High Performance Computing, Networking, ...

2020 arXiv
[38]

Florian Schneider and Sunayana Sitaram. 2024. https://aclanthology.org/2024.findings-emnlp.250 M5 -- a diverse benchmark to assess the performance of large multimodal models across multilingual and multicultural vision-language tasks . In Findings of the Association for Comput...

2024
[39]

Hao Shao, Shengju Qian, Han Xiao, Guanglu Song, Zhuofan Zong, Letian Wang, Yu Liu, and Hongsheng Li. 2024. https://arxiv.org/abs/2403.16999 Visual cot: Unleashing chain-of-thought reasoning in multi-modal language models . Preprint, arXiv:2403.16999

2024 arXiv
[40]

Hai-Long Sun, Da-Wei Zhou, Yang Li, Shiyin Lu, Chao Yi, Qing-Guo Chen, Zhao Xu, Weihua Luo, Kaifu Zhang, De-Chuan Zhan, et al. 2025. Parrot: Multilingual visual instruction tuning. In ICML

2025
[41]

Tianyi Tang, Wenyang Luo, Haoyang Huang, Dongdong Zhang, Xiaolei Wang, Xin Zhao, Furu Wei, and Ji-Rong Wen. 2024. https://api.semanticscholar.org/CorpusID:268032136 Language-specific neurons: The key to multilingual capabilities in large language models . In Annual Meeting of ...

2024
[42]

Chris Wendler, Veniamin Veselovsky, Giovanni Monea, and Robert West. 2024. https://doi.org/10.48550/ARXIV.2402.10588 Do llamas work in english? on the latent language of multilingual transformers . CoRR, abs/2402.10588

2024 doi arXiv
[43]

Shaolei Zhang, Qingkai Fang, Zhe Yang, and Yang Feng. 2025 a . https://api.semanticscholar.org/CorpusID:275342951 Llava-mini: Efficient image and video large multimodal models with one vision token . In Proceedings of XYZ Conference

2025
[44]

Xiaofeng Zhang, Yihao Quan, Chen Shen, Xiaosong Yuan, Shaotian Yan, Liang Xie, Wenxiao Wang, Chaochen Gu, Hao Tang, and Jieping Ye. 2025 b . From redundancy to relevance: Enhancing explainability in multimodal large language models. Annual Conference of the Nations of the Amer...

2025
[45]

Xueyan Zhang, Jinman Zhao, Zhifei Yang, Yibo Zhong, Shuhao Guan, Linbo Cao, and Yining Wang. 2025 c . https://aclanthology.org/2025.acl-long.575/ UORA : Uniform orthogonal reinitialization adaptation in parameter efficient fine-tuning of large models . In Proceedings of the 63...

2025
[46]

Zhi Zhang, Srishti Yadav, Fengze Han, and Ekaterina Shutova. 2024 a . https://doi.org/10.48550/ARXIV.2411.18620 Cross-modal information flow in multimodal large language models . CoRR, abs/2411.18620

2024 doi arXiv
[47]

Zhuosheng Zhang, Aston Zhang, Mu Li, Hai Zhao, George Karypis, and Alex Smola. 2024 b . https://openreview.net/forum?id=y1pPWFVfvR Multimodal chain-of-thought reasoning in language models . Trans. Mach. Learn. Res., 2024

2024
[48]

Jinman Zhao and Xueyan Zhang. 2024. https://openreview.net/forum?id=wLQ3I0F1oj Large language model is not a (multilingual) compositional relation reasoner . In First Conference on Language Modeling

2024
[49]

Yiran Zhao, Wenxuan Zhang, Guizhen Chen, Kenji Kawaguchi, and Lidong Bing. 2024. https://openreview.net/forum?id=ctXYOoAgRy How do large language models handle multilingualism? In The Thirty-eighth Annual Conference on Neural Information Processing Systems

2024
[50]

Yibo Zhong, Jinman Zhao, and Yao Zhou. 2025. https://aclanthology.org/2025.findings-acl.874/ Low-rank interconnected adaptation across layers . In Findings of the Association for Computational Linguistics, ACL 2025, Vienna, Austria, July 27 - August 1, 2025 , pages 17005--1702...

2025
[51]

Wenhao Zhu, Shujian Huang, Fei Yuan, Shuaijie She, Jiajun Chen, and Alexandra Birch. 2024. https://doi.org/10.18653/V1/2024.FINDINGS-ACL.498 Question translation training for better multilingual reasoning . In Findings of the Association for Computational Linguistics, ACL 2024...

2024 doi
[52]

online" 'onlinestring :=

ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...
[53]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [1]

Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, et al. 2023. Qwen technical report. arXiv preprint arXiv:2309.16609

arXiv 2023

[2] [2]

Jinze Bai, Shuai Bai, Shusheng Yang, Shijie Wang, Sinan Tan, Peng Wang, Junyang Lin, Chang Zhou, and Jingren Zhou. 2024. https://openreview.net/forum?id=qrGjFJVl3m Qwen- VL : A versatile vision-language model for understanding, localization, text reading, and beyond

work page 2024

[3] [3]

Samyadeep Basu, Martin Grayson, Cecily Morrison, Besmira Nushi, Soheil Feizi, and Daniela Massiceti. 2024. https://openreview.net/forum?id=s63dtq0mwA Understanding information storage and transfer in multi-modal large language models . In The Thirty-eighth Annual Conference on Neural Information Processing Systems

2024

[4] [4]

Thapliyal, Idan Szpektor, Julien Amelot, Xi Chen, and Radu Soricut

Soravit Changpinyo, Linting Xue, Michal Yarom, Ashish V. Thapliyal, Idan Szpektor, Julien Amelot, Xi Chen, and Radu Soricut. 2023. MaXM : Towards multilingual visual question answering. In Findings of the Association for Computational Linguistics: EMNLP

work page 2023

[5] [5]

Lin Chen, Jinsong Li, Xiaoyi Dong, Pan Zhang, Conghui He, Jiaqi Wang, Feng Zhao, and Dahua Lin. 2024. Sharegpt4v: Improving large multi-modal models with better captions. In European Conference on Computer Vision, pages 370--387. Springer

2024

[6] [6]

Xi Chen, Xiao Wang, Soravit Changpinyo, A. J. Piergiovanni, Piotr Padlewski, Daniel Salz, Sebastian Goodman, Adam Grycner, Basil Mustafa, Lucas Beyer, Alexander Kolesnikov, Joan Puigcerver, Nan Ding, Keran Rong, Hassan Akbari, Gaurav Mishra, Linting Xue, Ashish V. Thapliyal, James Bradbury, and Weicheng Kuo. 2023 a . https://openreview.net/forum?id=mWVoBz...

work page 2023

[7] [7]

Xi Chen, Xiao Wang, Soravit Changpinyo, A. J. Piergiovanni, Piotr Padlewski, Daniel Salz, Sebastian Goodman, Adam Grycner, Basil Mustafa, Lucas Beyer, Alexander Kolesnikov, Joan Puigcerver, Nan Ding, Keran Rong, Hassan Akbari, Gaurav Mishra, Linting Xue, Ashish V. Thapliyal, James Bradbury, and Weicheng Kuo. 2023 b . https://openreview.net/forum?id=mWVoBz...

work page 2023

[8] [8]

Gonzalez, Ion Stoica, and Eric P

Wei-Lin Chiang, Zhuohan Li, Zi Lin, Ying Sheng, Zhanghao Wu, Hao Zhang, Lianmin Zheng, Siyuan Zhuang, Yonghao Zhuang, Joseph E. Gonzalez, Ion Stoica, and Eric P. Xing. 2023. https://lmsys.org/blog/2023-03-30-vicuna/ Vicuna: An open-source chatbot impressing gpt-4 with 90\

2023