🚀 About Me
📫 E-mail : [email protected]
📥 Github : 🖱Github
💻 Velog : 📎Velog
📞 Phone : 010-4909-4946
💡 Strength:
- Proficient in analyzing code and comfortable with dissecting and modifying internal code of libraries like HuggingFace, Accelerate(fsdp), Django, etc.
- Quick at implementing code based on understanding of its logic.
- Possesses knowledge not only in AI but also across various areas of Computer Science.
- Enjoys engaging in conversations with others and is highly active in collaboration and communication.
📰 News
🔬 Publications
I0T: Embedding Standardization Method Toward Zero Modality Gap
- Na Min An*, Eunki Kim*, James Thorne, Hyunjung Shim
- ACL ‘25 Main
- TLDR: We attribute and mitigate explicit factors in the image and text embeddings that lead to the modality gap.
- arxiv: https://arxiv.org/abs/2412.14384
Learning to Insert [PAUSE] Tokens for Better Reasoning
- Eunki Kim*, Sangryul Kim*, James Thorne
- ACL ‘25 Findings
- TLDR: We introduce a novel approach termed Dynamic Inserting Token Training (DIT), which identifies positions within sequences where model confidence, determined by token log likelihood, is lowest, and strategically inserts [PAUSE] tokens to bolster the model's predictive capabilities for subsequent tokens.
- arxiv: https://arxiv.org/abs/2506.03616