44 citations · 46 across the 3 of their papers we have counts for
7 papers
Efficient Test-time Adaptation through Candidate Verification and Divergence Shifts
Seungmin Oh, Seunghun Kang, Jongbin Ryu
Vision-language models (VLMs) achieve strong zero-shot transferability but remain vulnerable to target-domain shifts at inference time. Test-time adaptation (TTA) offers a practica…
Differentiable Bit-Widths: Co-optimizing Pruning and Quantization via SVD for Ultra-Efficient LLM Compression
Hankyul Kang, Jongbin Ryu
SVD-based pruning and quantization have recently emerged as a promising strategy for the ultra-efficient compression of large language models. In these methods, compression is perf…
Layer-wise Curriculum Learning for Efficient LLM Compression
Donggeon Lee, Dooyeon Na, Seungmin Oh +1
In this paper, we introduce layer-wise curriculum learning for efficient LLM compression. The proposed method facilitates the knowledge transfer from the teacher model to the stude…
Colla-Q: Toward Collaborative Experts in MoE Quantization via Minimax Precision Balancing
Eunju Shin, Jongbin Ryu
In this paper, we present a Mixture-of-Experts (MoE) quantization method based on activation entropy. Although quantization reduces memory and computational costs, it can substanti…
Interactive Multi-Head Self-Attention with Linear Complexity
Hankyul Kang, Ming-Hsuan Yang, Jongbin Ryu
We propose an efficient interactive method for multi-head self-attention via decomposition. For existing methods using multi-head self-attention, the attention operation of each he…
Towards long-tailed, multi-label disease classification from chest X-ray: Overview of the CXR-LT challenge
Gregory Holste, Yiliang Zhou, Song Wang +22
Many real-world image recognition problems, such as diagnostic medical imaging exams, are "long-tailed" $\unicode{x2013}$ there are a few common findings followed by many more rela…
Robust Asymmetric Loss for Multi-Label Long-Tailed Learning
Wongi Park, Inhyuk Park, Sungeun Kim +1
In real medical data, training samples typically show long-tailed distributions with multiple labels. Class distribution of the medical data has a long-tailed shape, in which the i…