3 papers
cs.CL2025
Ko-MuSR: A Multistep Soft Reasoning Benchmark for LLMs Capable of Understanding Korean
Chanwoo Park, Suyoung Park, JiA Kang +6
We present Ko-MuSR, the first benchmark to comprehensively evaluate multistep, soft reasoning in long Korean narratives while minimizing data contamination. Built following MuSR, K…
cs.CL2025
Beyond Line-Level Filtering for the Pretraining Corpora of LLMs
Chanwoo Park, Suyoung Park, Yelim Ahn +3
While traditional line-level filtering techniques, such as line-level deduplication and trailing-punctuation filters, are commonly used, these basic methods can sometimes discard v…
cs.AI2025
PUZZLED: Jailbreaking LLMs through Word-Based Puzzles
Yelim Ahn, Jaejin Lee
As large language models (LLMs) are increasingly deployed across diverse domains, ensuring their safety has become a critical concern. In response, studies on jailbreak attacks hav…