Výsledky vyhledávání - "Kim, JinSeok"

Report

AMXFP4: Taming Activation Outliers with Asymmetric Microscaling Floating-Point for 4-bit LLM Inference

Autor: Lee, Janghwan, Park, Jiwoong, Kim, Jinseok, Kim, Yongjik, Oh, Jungju, Oh, Jinwook, Choi, Jungwook

Scaling Large Language Models (LLMs) with extended context lengths has increased the need for efficient low-bit quantization to manage their substantial computational demands. However, reducing precision to 4 bits frequently degrades performance due

Externí odkaz: http://arxiv.org/abs/2411.09909

Zobrazit plný text záznamu

Report

Safe-Embed: Unveiling the Safety-Critical Knowledge of Sentence Encoders

Autor: Kim, Jinseok, Jung, Jaewon, Kim, Sangyeop, Park, Sohyung, Cho, Sungzoon

Despite the impressive capabilities of Large Language Models (LLMs) in various tasks, their vulnerability to unsafe prompts remains a critical issue. These prompts can lead LLMs to generate responses on illegal or sensitive topics, posing a significa

Externí odkaz: http://arxiv.org/abs/2407.06851

Zobrazit plný text záznamu

Report

Beyond Binary Gender Labels: Revealing Gender Biases in LLMs through Gender-Neutral Name Predictions

Autor: You, Zhiwen, Lee, HaeJin, Mishra, Shubhanshu, Jeoung, Sullam, Mishra, Apratim, Kim, Jinseok, Diesner, Jana

Name-based gender prediction has traditionally categorized individuals as either female or male based on their names, using a binary classification system. That binary approach can be problematic in the cases of gender-neutral names that do not align

Externí odkaz: http://arxiv.org/abs/2407.05271

Zobrazit plný text záznamu

Report

Arbitrary-Scale Image Generation and Upsampling using Latent Diffusion Model and Implicit Neural Decoder

Autor: Kim, Jinseok, Kim, Tae-Kyun

Super-resolution (SR) and image generation are important tasks in computer vision and are widely adopted in real-world applications. Most existing methods, however, generate images only at fixed-scale magnification and suffer from over-smoothing and

Externí odkaz: http://arxiv.org/abs/2403.10255

Zobrazit plný text záznamu

Akademický článek

The Impact of Name Ambiguity on Properties of Coauthorship Networks

Autor: Kim, Jinseok, Kim, Heejun, Diesner, Jana

Publikováno v: Journal of Information Science Theory and Practice, Vol 2, Iss 2, Pp 6-15 (2014)

Initial based disambiguation of author names is a common data pre-processing step in bibliometrics. It is widely accepted that this procedure can introduce errors into network data and any subsequent analytical results. What is not sufficiently under

Externí odkaz: https://doaj.org/article/a5a553a6718442f59c909d0c1f45a95b

Zobrazit plný text záznamu

Vyhledávací nástroje:

Upřesnit hledání