Latticethe structure behind the noise

Papers Digest Topics Selected Labs Collections FAQ

Created by Flynn Lachendro

Papers Digest Topics Labs Saved

Search

Search papers, labs, and topics across Lattice.

Built by Flynn Lachendro·𝕏 / Twitter·RSS··FAQ·Glossary·Privacy

Jeonghoon Kim | Lattice

Jeonghoon Kim

NAVER Cloud, KAIST

Papers on Lattice

2

Total citations

0

Topics

4

Publication activitypapers/week, last 8 weeks

Research focus

Eval Frameworks & Benchmarks (2)Speech & Audio (1)Natural Language Processing (1)Training Efficiency & Optimization (1)

Frequent co-authors

Joosung Lee (2)Hwiyeol Jo (2)Jisu Jeon (1)Seungyeon Jwa (1)

Papers (2)

Jun 23, 2026

Hongik University2w ago·also KAIST, NAVER Labs, SNU

ParaPairAudioBench: Paralinguistic Pairwise Audio Benchmark for LALM-as-a-Judge

LALMs are falling short in paralinguistic evaluations, missing the mark by 32 percentage points compared to human judges.

Jisu Jeon, Seungyeon Jwa, Joosung Lee +5

Eval Frameworks & Benchmarks Speech & Audio

Apr 7, 2026

Apr 7, 2026·also KAIST, SNU

What Models Know, How Well They Know It: Knowledge-Weighted Fine-Tuning for Learning When to Say "I Don't Know"

LLMs can learn to say "I don't know" more effectively by weighting the fine-tuning signal based on how well they already know the answer.

Joosung Lee, Hwiyeol Jo, Donghyeon Ko +3

Eval Frameworks & Benchmarks Natural Language Processing Training Efficiency & Optimization

Jinhyeon Kim (1)

Woojin Chung (1)

Donghyeon Ko (1)