Seungyoo Lee
  • Home
  • Publications
  • Experience
  • Teaching & Talks
  • CV
  • Publications
    • Mixture-Trained Merging for Unified Multi-Objective Models
    • SAVE: Sparsity-Aware Influence Estimation for Vocabulary-Expanded LLMs
    • Confidence is Not Universal: Task-Dependent Calibration and Emergent Behavior in LLMs
    • From Drift to Coherence: Stabilizing Beliefs in LLMs
    • OCNR: Stabilizing Self-Play by Mitigating Iteration-Collapse With One-Class Novelty Rewards
    • Deep Learning-Based Liver Tumor Segmentation from Computed Tomography Scans with a Gradient-Enhanced Network
    • Automatic Liver Couinaud Segmentation from Computed Tomography Scans with a Gradient Enhanced Hierarchical Cascade Deep Learning Network
    • PANGEA: Projection-Based Augmentation with Non-Relevant General Data for Enhanced Domain Adaptation in LLMs
    • Variational Bayesian Pseudo-Coreset
    • G-UNETR++: A Gradient-Enhanced Network for Accurate and Robust Liver Segmentation from Computed Tomography Images
    • Estimation of Right Lobe Graft Weight for Living Donor Liver Transplantation Using Deep Learning-Based Fully Automatic Computed Tomographic Volumetry
    • Human Activity Recognition using Sensor Series Transformer with High-frequency Positional Encoding
    • Vehicle Distance Estimation from a Monocular Camera for Advanced Driver Assistance Systems
  • Experience
  • Teaching & Talks

Mixture-Trained Merging for Unified Multi-Objective Models

Sep 24, 2026·
SeongHyeon Kim*
,
Chaeyun Jang*
Seungyoo Lee
Seungyoo Lee
,
Jiyeon Ham
,
Yunju Bak
,
Boseop Kim
,
Juho Lee
· 0 min read
Source Document arXiv
Type
Conference paper
Publication
In The Fortieth Annual Conference on Neural Information Processing Systems (NeurIPS 2026)
Accepted
September 24, 2026
Last updated on Sep 24, 2026
Seungyoo Lee
Authors
Seungyoo Lee
Ph.D. Student in AI

SAVE: Sparsity-Aware Influence Estimation for Vocabulary-Expanded LLMs Sep 24, 2026 →

© 2026 Seungyoo Lee · Last updated November 2026

Built with Hugo Blox