Hyeonbin Hwang

@ronalhwang

Ph.D Student @KAIST_AI - 𝗥𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 / 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗦𝘆𝘀𝘁𝗲𝗺𝘀

Seoul, South Korea

hbin0701.github.io

Joined February 2023

407 Following

466 Followers

Replies Reposts Media

Hyeonbin Hwang
@ronalhwang
Sep 28
Last year, I had the honor to contribute to one of the most exciting projects in my life! A gaming agent that can play with you real time! 👾 Check out the tech report for more details!
Hyunseung Kim
@mynsng_
Sep 28
Excited to share the PUBG Ally Tech Report! 📄 arxiv.org/abs/2609.29837 In June, we introduced PUBG Ally, the world’s first in-game AI agent. Check out the unique challenges behind its architecture, data, training, evaluation, and on-device optimization!
Hyeonbin Hwang
@ronalhwang
Oct 7, 2024
New Paper Alert📢 Q: Is more pre-training always better? 🤔 Not always—your LLM might lose its plasticity 🧠 We introduce "Knowledge Entropy" 📊, change in the knowledge acquisition and retention ability during pretraining. Check out the paper for more details! 😃
Jiyeon Kim
@jiyeonkimd
Oct 7, 2024
❓Do LLMs maintain the capability of knowledge acquisition throughout pretraining? If not, what is driving force behind it? ❗Our findings reveal that decreasing knowledge entropy hinders knowledge acquisition and retention as pretraining progresses. 📄arxiv.org/abs/2410.01380
Hyeonbin Hwang
@ronalhwang
Jun 12, 2024
I have contributed some reasoning tasks here, and made sure they cannot be easily solved simply because LLMs could have seen similar instances! While many were skeptical of LLM evaluation benchmarks these days, I think this is a really valuable asset. Give it a try! :)
Seungone Kim
@seungonekim
Jun 12, 2024
🤔How can we systematically assess an LM's proficiency in a specific capability without using summary measures like helpfulness or simple proxy tasks like multiple-choice QA? Introducing the ✨BiGGen Bench, a benchmark that directly evaluates nine core capabilities of LMs.
Hyeonbin Hwang
@ronalhwang
Feb 3, 2024
Does anyone know whether future tokens affect the result of previous tokens’ logits? I thought they didn’t.. but my results say otherwise... I'm using Mistral-7b. I posted the question at stack-overflow for more details: stackoverflow.com/questions/7793…

Hyeonbin Hwang

Hyeonbin Hwang’s posts