OSI LAB
.
people
publications
projects
news
apply
π·
photos
π§βπ»
members
News
[Oct 25, 2024]
1 paper accepted to TMLR!
FLR: Label-Mixture Regularization for Federated Learning with Noisy Labels
[Sep 26, 2024]
8 papers accepted to NeurIPS 2024 (6 main, 2 workshop)
Main:
Conditional Synthesis of 3D Molecules with Time Correction Sampler
An Adaptive Approach for Infinitely Many-armed Bandits under Generalized Rotting Constraints
A Unified Confidence Sequence for Generalized Linear Models, with Applications to Bandits
Preference Alignment with Flow Matching
Block Transformer: Global-to-Local Language Modeling for Fast Inference
Exploring & Improving Multi-token Prediction (Block Draft) in Language Modeling
Workshop:
A Unified Framework for Speculative Decoding with Multiple Drafters as a Bandit
(
ENSLP Workshop
)
FlickerFusion: Intra-trajectory Domain Generalizing Multi-agent Reinforcement Learning
(
OWA Workshop
)
[Sep 20, 2024]
5 papers accepted to EMNLP 2024 (4 main, 1 finding)
Main:
Towards Fast Multilingual LLM Inference: Speculative Decoding and Specialized Drafters
Stable Language Model Pre-training by Reducing Embedding Variability
DocKD: Knowledge Distillation from LLMs for Open-World Document Understanding Models
Towards Difficulty-Agnostic Efficient Transfer Learning for Vision-Language Models
Findings:
BAPO: Base-Anchored Preference Optimization for Personalized Alignment in Large Language Models
[Sep 05, 2024]
1 paper accepted to TMLR!
Non-backtracking Graph Neural Networks
[Jul 30, 2024]
1 paper at CKAIA 2024 (νκ΅μΈκ³΅μ§λ₯νν) selected as one of the *best papers*
A Unified Confidence Sequence for Generalized Linear Models, with Applications to Bandits
[May 04, 2024]
2 papers accepted to ICML 2024!
DistiLLM: Towards Streamlined Distillation for Large Language Models
Synergistic Integration of Coordinate Network and Tensorial Feature for Improving Neural Radiance Fields from Sparse Inputs
[Apr 13, 2024]
1 paper at ICASSP 2024 selected as the *best student paper*
STaR: Distilling Speech Temporal Relation for Lightweight Speech Self-Supervised Learning Models
[Mar 15, 2024]
1 paper accepted to NAACL 2024!
Carpe diem: On the Evaluation of World Knowledge in Lifelong Language Models
[Feb 28, 2024]
1 paper accepted to CVPR 2024!
FedSOL: Stabilized Orthogonal Learning in Federated Learning
[Jan 20, 2024]
1 paper accepted to AISTATS 2024.
Improved Regret Bounds of (Multinomial) Logistic Bandits via Regret-to-Confidence-Set Conversion
[Jan 16, 2024]
2 papers accepted to ICLR 2024 (1 spotlight).
Querying Easily Flip-flopped Samples for Deep Active Learning
Instructive Decoding: Instruction-Tuned Large Language Models are Self-Refiner from Noisy Instructions
(spotlight)
[Dec 13, 2023]
2 papers accepted to ICASSP 2024.
Stethoscope-guided Supervised Contrastive Learning for Cross-domain Adaptation on Respiratory Sound Classification
STaR: Distilling Speech Temporal Relation for Lightweight Speech Self-Supervised Learning Models
[Dec 09, 2023]
1 paper accepted to AAAI 2024!
Leveraging Normalization Layer in Adapters With Progressive Learning and Adaptive Distillation for Cross-Domain Few-Shot Learning
[Nov 23, 2023]
1 paper at OPODIS 2023 selected as the *best student paper*
Flooding with Absorption: An Efficient Protocol for Heterogeneous Bandits over Complex Networks
[Oct 30, 2023]
2 papers accepted to OPODIS 2023
Flooding with Absorption: An Efficient Protocol for Heterogeneous Bandits over Complex Networks
On the convergence time in graphical games: a locality-sensitive approach
[Oct 09, 2023]
4 papers accepted to EMNLP 2023 (1 main, 3 finding)
Fast and Robust Early-Exiting Framework for Autoregressive Language Models with Synchronized Parallel Decoding
HARE: Explainable Hate Speech Detection with Step-by-Step Reasoning
(Findings)
NASH: A Simple Unified Framework of Structured Pruning for Accelerating Encoder-Decoder Language Models
(Findings)
Bayesian Multi-Task Transfer Learning for Soft Prompt Tuning
(Findings)
[Sep 21, 2023]
3 papers accepted to NeurIPS 2023
Navigating Data Heterogeneity in Federated Learning: A Semi-Supervised Approach for Object Detection
Fair Streaming Principal Component Analysis: Statistical and Algorithmic Viewpoint
PLASTIC: Improving Input and Label Plasticity for Sample Efficient Reinforcement Learning
[May 18, 2023]
2 papers accepted to INTERSPEECH 2023.
Patch-Mix Contrastive Learning with Audio Spectrogram Transformer on Respiratory Sound Classification
(
media
)
Recycle-and-Distill: Universal Compression Strategy for Transformer-based Speech SSL Models with Attention Map Reusing and Masking Distillation
[May 01, 2023]
1 paper accepted to ACL 2023.
Large Language Models Are Reasoning Teachers
[Feb 28, 2023]
2 papers accepted to CVPR 2023.
Coreset Sampling from Open-Set for Fine-Grained Self-Supervised Learning
Re-thinking Federated Active Learning based on Inter-class Diversity
[Jan 20, 2023]
1 paper accepted to EACL 2023 (findings).
Revisiting Intermediate Layer Distillation for Compressing Language Models: An Overfitting Perspective
[Jan 20, 2023]
2 papers accepted to ICLR 2023 (1 spotlight).
CUDA: Curriculum of Data Augmentation for Long-tailed Recognition
(spotlight)
Mitigating Dataset Bias by Using Per-Sample Gradient
[Jan 19, 2023]
2 papers accepted to AISTATS 2023.
Contextual Linear Bandits under Noisy Features: Towards Bayesian Oracles
Nearly Optimal Latent State Decoding in Block MDPs
[Jan 04, 2023]
1 paper accepted to AAMAS 2023.
Toward Risk-based Optimistic Exploration for Cooperative Multi-Agent Reinforcement Learning
[Dec 07, 2022]
OSI LAB wins 1st place in the Cell Segmentation Challenge at NeurIPS 2022.
Team: Gihun Lee*, Sangmook Kim*, Joonkee Kim*
MEDIAR: Harmony of Data-Centric and Model-Centric for Multi-Modality Microscopy
[Nov 18, 2022]
3 papers accepted to AAAI 2023 (2 orals).
Denoising After Entropy-based Debiasing A Robust Training Method for Dataset Bias with Noisy Labels
(oral)
Self-Contrastive Learning: Single-viewed Supervised Contrastive Framework using Sub-network
(oral)
A Gift from Label Smoothing: Robust Training with Adaptive Label Smoothing via Auxiliary Classifier under Label Noise
[Jul 28, 2022]
1 paper at CKAIA 2022 (νκ΅μΈκ³΅μ§λ₯νν) selected as one of the *best papers*
Near-Optimal Clustering in Block MDPs with Implications on Reward-Free RL
[Nov 02, 2021]
1 paper at JKAIA 2021 (νκ΅μΈκ³΅μ§λ₯νν) selected as one of the *best papers*, and 1 paper as one of the *honorable mentions*
best paper:
MMD-based Fair PCA via Manifold Optimization
honorable mention:
Self-Contrastive Learning