Search Results for author: Sang-Woo Lee

Found 32 papers, 18 papers with code

Attribute Injection for Pretrained Language Models: A New Benchmark and an Efficient Method

1 code implementation COLING 2022 Reinald Kim Amplayo, Kang Min Yoo, Sang-Woo Lee

Metadata attributes (e. g., user and product IDs from reviews) can be incorporated as additional inputs to neural-based NLP models, by expanding the architecture of the models to improve performance.

Attribute

Universal Domain Adaptation for Robust Handling of Distributional Shifts in NLP

1 code implementation23 Oct 2023 Hyuhng Joon Kim, Hyunsoo Cho, Sang-Woo Lee, Junyeob Kim, Choonghyun Park, Sang-goo Lee, Kang Min Yoo, Taeuk Kim

When deploying machine learning systems to the wild, it is highly desirable for them to effectively leverage prior knowledge to the unfamiliar domain while also firing alarms to anomalous inputs.

Universal Domain Adaptation

Query-Efficient Black-Box Red Teaming via Bayesian Optimization

1 code implementation27 May 2023 Deokjae Lee, JunYeong Lee, Jung-Woo Ha, Jin-Hwa Kim, Sang-Woo Lee, Hwaran Lee, Hyun Oh Song

To this end, we propose Bayesian red teaming (BRT), novel query-efficient black-box red teaming methods based on Bayesian optimization, which iteratively identify diverse positive test cases leading to model failures by utilizing the pre-defined user input pool and the past evaluations.

Bayesian Optimization Language Modelling

Asking Clarification Questions to Handle Ambiguity in Open-Domain QA

1 code implementation23 May 2023 Dongryeol Lee, Segwang Kim, Minwoo Lee, Hwanhee Lee, Joonsuk Park, Sang-Woo Lee, Kyomin Jung

We first present CAMBIGNQ, a dataset consisting of 5, 654 ambiguous questions, each with relevant passages, possible answers, and a clarification question.

Open-Domain Question Answering

Can Current Task-oriented Dialogue Models Automate Real-world Scenarios in the Wild?

no code implementations20 Dec 2022 Sang-Woo Lee, Sungdong Kim, Donghyeon Ko, Donghoon Ham, Youngki Hong, Shin Ah Oh, Hyunhoon Jung, Wangkyo Jung, Kyunghyun Cho, Donghyun Kwak, Hyungsuk Noh, WooMyoung Park

Task-oriented dialogue (TOD) systems are mainly based on the slot-filling-based TOD (SF-TOD) framework, in which dialogues are broken down into smaller, controllable units (i. e., slots) to fulfill a specific task.

Language Modelling Position +2

Keep Me Updated! Memory Management in Long-term Conversations

no code implementations17 Oct 2022 Sanghwan Bae, Donghyun Kwak, Soyoung Kang, Min Young Lee, Sungdong Kim, Yuin Jeong, Hyeri Kim, Sang-Woo Lee, WooMyoung Park, Nako Sung

Remembering important information from the past and continuing to talk about it in the present are crucial in long-term conversations.

Management

Continuous Decomposition of Granularity for Neural Paraphrase Generation

1 code implementation COLING 2022 Xiaodong Gu, Zhaowei Zhang, Sang-Woo Lee, Kang Min Yoo, Jung-Woo Ha

While Transformers have had significant success in paragraph generation, they treat sentences as linear sequences of tokens and often neglect their hierarchical information.

Paraphrase Generation Sentence

Leveraging Pre-Trained Language Models to Streamline Natural Language Interaction for Self-Tracking

no code implementations31 May 2022 Young-Ho Kim, Sungdong Kim, Minsuk Chang, Sang-Woo Lee

Current natural language interaction for self-tracking tools largely depends on bespoke implementation optimized for a specific tracking theme and data format, which is neither generalizable nor scalable to a tremendous design space of self-tracking.

Ground-Truth Labels Matter: A Deeper Look into Input-Label Demonstrations

no code implementations25 May 2022 Kang Min Yoo, Junyeob Kim, Hyuhng Joon Kim, Hyunsoo Cho, Hwiyeol Jo, Sang-Woo Lee, Sang-goo Lee, Taeuk Kim

Despite recent explosion of interests in in-context learning, the underlying mechanism and the precise impact of the quality of demonstrations remain elusive.

In-Context Learning Language Modelling

Mutual Information Divergence: A Unified Metric for Multimodal Generative Models

1 code implementation25 May 2022 Jin-Hwa Kim, Yunji Kim, Jiyoung Lee, Kang Min Yoo, Sang-Woo Lee

Based on a recent trend that multimodal generative evaluations exploit a vison-and-language pre-trained model, we propose the negative Gaussian cross-mutual information using the CLIP features as a unified metric, coined by Mutual Information Divergence (MID).

Hallucination Pair-wise Detection (1-ref) Hallucination Pair-wise Detection (4-ref) +5

Two-Step Question Retrieval for Open-Domain QA

1 code implementation Findings (ACL) 2022 Yeon Seonwoo, Juhee Son, Jiho Jin, Sang-Woo Lee, Ji-Hoon Kim, Jung-Woo Ha, Alice Oh

These models have shown a significant increase in inference speed, but at the cost of lower QA performance compared to the retriever-reader models.

Computational Efficiency Retrieval +1

Building a Role Specified Open-Domain Dialogue System Leveraging Large-Scale Language Models

1 code implementation NAACL 2022 Sanghwan Bae, Donghyun Kwak, Sungdong Kim, Donghoon Ham, Soyoung Kang, Sang-Woo Lee, WooMyoung Park

In this work, we study the challenge of imposing roles on open-domain dialogue systems, with the goal of making the systems maintain consistent roles while conversing naturally with humans.

Few-Shot Learning

Plug-and-Play Adaptation for Continuously-updated QA

no code implementations Findings (ACL) 2022 Kyungjae Lee, Wookje Han, Seung-won Hwang, Hwaran Lee, Joonsuk Park, Sang-Woo Lee

To this end, we first propose a novel task--Continuously-updated QA (CuQA)--in which multiple large-scale updates are made to LMs, and the performance is measured with respect to the success in adding and updating knowledge while retaining existing knowledge.

Efficient Attribute Injection for Pretrained Language Models

no code implementations16 Sep 2021 Reinald Kim Amplayo, Kang Min Yoo, Sang-Woo Lee

Metadata attributes (e. g., user and product IDs from reviews) can be incorporated as additional inputs to neural-based NLP models, by modifying the architecture of the models, in order to improve their performance.

Attribute

Weakly Supervised Pre-Training for Multi-Hop Retriever

1 code implementation Findings (ACL) 2021 Yeon Seonwoo, Sang-Woo Lee, Ji-Hoon Kim, Jung-Woo Ha, Alice Oh

In multi-hop QA, answering complex questions entails iterative document retrieval for finding the missing entity of the question.

Retrieval

Efficient Dialogue State Tracking by Selectively Overwriting Memory

3 code implementations ACL 2020 Sungdong Kim, Sohee Yang, Gyuwan Kim, Sang-Woo Lee

This mechanism consists of two steps: (1) predicting state operation on each of the memory slots, and (2) overwriting the memory with new values, of which only a few are generated according to the predicted state operations.

Dialogue State Tracking Multi-domain Dialogue State Tracking

Large-Scale Answerer in Questioner's Mind for Visual Dialog Question Generation

1 code implementation ICLR 2019 Sang-Woo Lee, Tong Gao, Sohee Yang, Jaejun Yoo, Jung-Woo Ha

Answerer in Questioner's Mind (AQM) is an information-theoretic framework that has been recently proposed for task-oriented dialog systems.

Question Generation Question-Generation +1

Answerer in Questioner's Mind: Information Theoretic Approach to Goal-Oriented Visual Dialog

1 code implementation NeurIPS 2018 Sang-Woo Lee, Yu-Jung Heo, Byoung-Tak Zhang

Goal-oriented dialogue tasks occur when a questioner asks an action-oriented question and an answerer responds with the intent of letting the questioner know a correct action to take.

Goal-Oriented Dialog Visual Dialog

Overcoming Catastrophic Forgetting by Incremental Moment Matching

1 code implementation NeurIPS 2017 Sang-Woo Lee, Jin-Hwa Kim, Jaehyun Jun, Jung-Woo Ha, Byoung-Tak Zhang

Catastrophic forgetting is a problem of neural networks that loses the information of the first task after training the second task.

Transfer Learning

Dual Memory Architectures for Fast Deep Learning of Stream Data via an Online-Incremental-Transfer Strategy

no code implementations15 Jun 2015 Sang-Woo Lee, Min-Oh Heo, Jiwon Kim, Jeonghee Kim, Byoung-Tak Zhang

The proposed architecture consists of deep representation learners and fast learnable shallow kernel networks, both of which synergize to track the information of new data.

Transfer Learning

Cannot find the paper you are looking for? You can Submit a new open access paper.