Výsledky vyhledávání - "Kim, Junyeob"

Report

Adaptive Contrastive Decoding in Retrieval-Augmented Generation for Handling Noisy Contexts

Autor: Kim, Youna, Kim, Hyuhng Joon, Park, Cheonbok, Park, Choonghyun, Cho, Hyunsoo, Kim, Junyeob, Yoo, Kang Min, Lee, Sang-goo, Kim, Taeuk

When using large language models (LLMs) in knowledge-intensive tasks, such as open-domain question answering, external context can bridge the gap between external knowledge and the LLMs' parametric knowledge. Recent research has been developed to amp

Externí odkaz: http://arxiv.org/abs/2408.01084

Zobrazit plný text záznamu

Report

Investigating the Influence of Prompt-Specific Shortcuts in AI Generated Text Detection

Autor: Park, Choonghyun, Kim, Hyuhng Joon, Kim, Junyeob, Kim, Youna, Kim, Taeuk, Cho, Hyunsoo, Jo, Hwiyeol, Lee, Sang-goo, Yoo, Kang Min

AI Generated Text (AIGT) detectors are developed with texts from humans and LLMs of common tasks. Despite the diversity of plausible prompt choices, these datasets are generally constructed with a limited number of prompts. The lack of prompt variati

Externí odkaz: http://arxiv.org/abs/2406.16275

Zobrazit plný text záznamu

Report

Aligning Language Models to Explicitly Handle Ambiguity

Autor: Kim, Hyuhng Joon, Kim, Youna, Park, Cheonbok, Kim, Junyeob, Park, Choonghyun, Yoo, Kang Min, Lee, Sang-goo, Kim, Taeuk

In interactions between users and language model agents, user utterances frequently exhibit ellipsis (omission of words or phrases) or imprecision (lack of exactness) to prioritize efficiency. This can lead to varying interpretations of the same inpu

Externí odkaz: http://arxiv.org/abs/2404.11972

Zobrazit plný text záznamu

Report

Universal Domain Adaptation for Robust Handling of Distributional Shifts in NLP

Autor: Kim, Hyuhng Joon, Cho, Hyunsoo, Lee, Sang-Woo, Kim, Junyeob, Park, Choonghyun, Lee, Sang-goo, Yoo, Kang Min, Kim, Taeuk

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. In order to address these requirements, Universal Doma

Externí odkaz: http://arxiv.org/abs/2310.14849

Zobrazit plný text záznamu

Report

Prompt-Augmented Linear Probing: Scaling beyond the Limit of Few-shot In-Context Learners

Autor: Cho, Hyunsoo, Kim, Hyuhng Joon, Kim, Junyeob, Lee, Sang-Woo, Lee, Sang-goo, Yoo, Kang Min, Kim, Taeuk

Through in-context learning (ICL), large-scale language models are effective few-shot learners without additional model fine-tuning. However, the ICL performance does not scale well with the number of available training samples as it is limited by th

Externí odkaz: http://arxiv.org/abs/2212.10873

Zobrazit plný text záznamu

Report

Self-Generated In-Context Learning: Leveraging Auto-regressive Language Models as a Demonstration Generator

Autor: Kim, Hyuhng Joon, Cho, Hyunsoo, Kim, Junyeob, Kim, Taeuk, Yoo, Kang Min, Lee, Sang-goo

Large-scale pre-trained language models (PLMs) are well-known for being capable of solving a task simply by conditioning a few input-label pairs dubbed demonstrations on a prompt without being explicitly tuned for the desired downstream task. Such a

Externí odkaz: http://arxiv.org/abs/2206.08082

Zobrazit plný text záznamu

Report

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

Autor: Yoo, Kang Min, Kim, Junyeob, Kim, Hyuhng Joon, Cho, Hyunsoo, Jo, Hwiyeol, Lee, Sang-Woo, Lee, Sang-goo, Kim, Taeuk

Despite recent explosion of interests in in-context learning, the underlying mechanism and the precise impact of the quality of demonstrations remain elusive. Intuitively, ground-truth labels should have as much impact in in-context learning (ICL) as

Externí odkaz: http://arxiv.org/abs/2205.12685

Zobrazit plný text záznamu