Jiaang Li
@jiaangli.bsky.social
PhD student at University of Copenhagen @belongielab.org | #nlp #computervision | ELLIS student @ellis.eu
🌐 https://jiaangli.github.io/
🌐 https://jiaangli.github.io/
Great collaboration with @yfyuan01.bsky.social @wenyan62.bsky.social @aliannejadi.bsky.social @danielhers.bsky.social , Anders Søgaard, Ivan Vulić, Wenxuan Zhang, Paul Liang, Yang Deng, @serge.belongie.com
May 23, 2025 at 5:04 PM
Great collaboration with @yfyuan01.bsky.social @wenyan62.bsky.social @aliannejadi.bsky.social @danielhers.bsky.social , Anders Søgaard, Ivan Vulić, Wenxuan Zhang, Paul Liang, Yang Deng, @serge.belongie.com
🔗More here:
Project Page: jiaangli.github.io/RAVENEA/
Code: github.com/yfyuan01/RAV...
Dataset: huggingface.co/datasets/jaa...
Project Page: jiaangli.github.io/RAVENEA/
Code: github.com/yfyuan01/RAV...
Dataset: huggingface.co/datasets/jaa...
jaagli/ravenea · Datasets at Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
May 23, 2025 at 5:04 PM
🔗More here:
Project Page: jiaangli.github.io/RAVENEA/
Code: github.com/yfyuan01/RAV...
Dataset: huggingface.co/datasets/jaa...
Project Page: jiaangli.github.io/RAVENEA/
Code: github.com/yfyuan01/RAV...
Dataset: huggingface.co/datasets/jaa...
📊Our experiments demonstrate that even lightweight VLMs, when augmented with culturally relevant retrievals, outperform their non-augmented counterparts and even surpass the next larger model tier, achieving at least a 3.2% improvement in cVQA and 6.2% in cIC.
May 23, 2025 at 5:04 PM
📊Our experiments demonstrate that even lightweight VLMs, when augmented with culturally relevant retrievals, outperform their non-augmented counterparts and even surpass the next larger model tier, achieving at least a 3.2% improvement in cVQA and 6.2% in cIC.
🛠Culture-Aware Contrastive Learning
We propose Culture-aware Contrastive (CAC) Learning, a supervised learning framework compatible with both CLIP and SigLIP architectures. Fine-tuning with CAC can help models better capture culturally significant content.
We propose Culture-aware Contrastive (CAC) Learning, a supervised learning framework compatible with both CLIP and SigLIP architectures. Fine-tuning with CAC can help models better capture culturally significant content.
May 23, 2025 at 5:04 PM
🛠Culture-Aware Contrastive Learning
We propose Culture-aware Contrastive (CAC) Learning, a supervised learning framework compatible with both CLIP and SigLIP architectures. Fine-tuning with CAC can help models better capture culturally significant content.
We propose Culture-aware Contrastive (CAC) Learning, a supervised learning framework compatible with both CLIP and SigLIP architectures. Fine-tuning with CAC can help models better capture culturally significant content.
📚 Dataset Construction
RAVENEA integrates 1,800+ images, 2,000+ culture-related questions, 500+ human captions, and 10,000+ human-ranked Wikipedia documents to support two key tasks:
🎯Culture-focused Visual Question Answering (cVQA)
📝Culture-informed Image Captioning (cIC)
RAVENEA integrates 1,800+ images, 2,000+ culture-related questions, 500+ human captions, and 10,000+ human-ranked Wikipedia documents to support two key tasks:
🎯Culture-focused Visual Question Answering (cVQA)
📝Culture-informed Image Captioning (cIC)
May 23, 2025 at 5:04 PM
📚 Dataset Construction
RAVENEA integrates 1,800+ images, 2,000+ culture-related questions, 500+ human captions, and 10,000+ human-ranked Wikipedia documents to support two key tasks:
🎯Culture-focused Visual Question Answering (cVQA)
📝Culture-informed Image Captioning (cIC)
RAVENEA integrates 1,800+ images, 2,000+ culture-related questions, 500+ human captions, and 10,000+ human-ranked Wikipedia documents to support two key tasks:
🎯Culture-focused Visual Question Answering (cVQA)
📝Culture-informed Image Captioning (cIC)
Super cool! Incidentally, in our previous project, we also found that linear alignment between embedding spaces from two modalities is viable — and the alignment improves as LLMs scale.
bsky.app/profile/jiaa...
bsky.app/profile/jiaa...
🤔Do Vision and Language Models Share Concepts? 🚀
We present an empirical evaluation and find that language models partially converge towards representations isomorphic to those of vision models. #EMNLP
📃 direct.mit.edu/tacl/article...
We present an empirical evaluation and find that language models partially converge towards representations isomorphic to those of vision models. #EMNLP
📃 direct.mit.edu/tacl/article...
May 23, 2025 at 1:59 PM
Super cool! Incidentally, in our previous project, we also found that linear alignment between embedding spaces from two modalities is viable — and the alignment improves as LLMs scale.
bsky.app/profile/jiaa...
bsky.app/profile/jiaa...
Great collaboration with @constanzafierro.bsky.social , @YovaKem_v2, and Anders Søgaard!
👨💻 github.com/jiaangli/VLCA
📃 direct.mit.edu/tacl/article...
👨💻 github.com/jiaangli/VLCA
📃 direct.mit.edu/tacl/article...
GitHub - jiaangli/VLCA: Do Vision and Language Models Share Concepts? A Vector Space Alignment Study
Do Vision and Language Models Share Concepts? A Vector Space Alignment Study - jiaangli/VLCA
github.com
November 19, 2024 at 1:27 PM
Great collaboration with @constanzafierro.bsky.social , @YovaKem_v2, and Anders Søgaard!
👨💻 github.com/jiaangli/VLCA
📃 direct.mit.edu/tacl/article...
👨💻 github.com/jiaangli/VLCA
📃 direct.mit.edu/tacl/article...
🌱We then discuss the implications of our finding:
- the LM understanding debate
- the study of emergent properties
- philosophy
🧵(6/8)
- the LM understanding debate
- the study of emergent properties
- philosophy
🧵(6/8)
November 19, 2024 at 1:12 PM
🌱We then discuss the implications of our finding:
- the LM understanding debate
- the study of emergent properties
- philosophy
🧵(6/8)
- the LM understanding debate
- the study of emergent properties
- philosophy
🧵(6/8)