Papers by KyungTae Lim
Beyond Accuracy: Alignment and Error Detection across Languages in the Bi-GSM8K Math-Teaching Benchmark (2026.findings-eacl)
Copied to clipboard
| Challenge: | Recent advances in LLMs have significantly improved mathematical problem-solving, with models like GPT-4 achieving human-level performance. |
| Approach: | They propose a bilingual English-Korean dataset enriched with teacher solutions, student solutions, and annotations marking students’ initial errors. |
| Outcome: | The proposed model achieves high agreement with human judgments and lower latency and resource usage than commercial APIs, demonstrating strong computational efficiency. |
K-UniMorph: Korean Universal Morphology and its Feature Schema (2023.findings-acl)
Copied to clipboard
| Challenge: | Previously, the Korean language has been underrepresented in the field of morphological paradigms amongst hundreds of diverse world languages. |
| Approach: | They propose a new Universal Morphology dataset for Korean that preserves its distinct characteristics. |
| Outcome: | The proposed dataset extracts inflected Korean verb forms from the largest annotated corpus for Korean. |
TReX: Tokenizer Regression for Optimal Data Mixture (2026.eacl-long)
Copied to clipboard
| Challenge: | Existing approaches to train and inference tokenizers rely on heuristics or large-scale searches to determine optimal data mixtures. |
| Approach: | They propose a regression-based framework that efficiently predicts the optimal data mixture for tokenizer training. |
| Outcome: | The proposed model outperforms mixtures based on LLaMA3 and uniform distributions by up to 12% in both in- and out-of-distribution compression efficiency. |
X-LLaVA: Optimizing Bilingual Large Vision-Language Alignment (2024.findings-naacl)
Copied to clipboard
DongJae Shin, HyeonSeok Lim, Inho Won, ChangSu Choi, Minjun Kim, SeungWoo Song, HanGyeol Yoo, SangMin Kim, KyungTae Lim
| Challenge: | constructing multilingual data for large multimodal models presents its own set of challenges due to language diversity and complexity. |
| Approach: | They propose to use GPT4-V to construct multimodal training datasets using a text-only version of GPT4. |
| Outcome: | The proposed method performs well in Korean and English, surpassing existing methods. |
A Linguistically-Informed Annotation Strategy for Korean Semantic Role Labeling (2024.lrec-main)
Copied to clipboard
| Challenge: | Semantic role labeling is an essential component of semantic and syntactic processing of natural languages. |
| Approach: | They propose an annotation strategy for Korean semantic role labeling that is in line with the previously proposed linguistic theories as well as the distinct properties of the Korean language. |
| Outcome: | The proposed annotation strategy is consistent with the proposed linguistic theories and the distinct properties of the Korean language. |
Yet Another Format of Universal Dependencies for Korean (2022.coling-1)
Copied to clipboard
Yige Chen, Eunkyul Leah Jo, Yundong Yao, KyungTae Lim, Miikka Silfverberg, Francis M. Tyers, Jungyeul Park
| Challenge: | Existing dependency parsers for Korean do not perform as well as their English counterparts due to the complexity of Korean's linguistic features. |
| Approach: | They propose a morpheme-based Korean dependency parsing format and propose to adopt it to Universal Dependencies. |
| Outcome: | The proposed format outperforms parsing results for Korean UD treebanks and detailed error analysis. |
TELLME: Test-Enhanced Learning for Language Model Enrichment (2026.findings-eacl)
Copied to clipboard
Minjun Kim, Inho Won, HyeonSeok Lim, MinKyu Kim, Junghun Yuk, Wooyoung Go, Jongyoul Park, Jungyeul Park, KyungTae Lim
| Challenge: | Continual pre-training (CPT) has been widely adopted as a method for domain expansion in large language models, but has faced challenges such as acquiring large-scale domain-specific datasets and high computational costs. |
| Approach: | They propose a method that integrates the Test-Enhanced Learning principle with CPT to promote efficient domain-specific knowledge acquisition and long-term memory retention. |
| Outcome: | The proposed method outperforms existing methods by 23.6% in the financial domain and achieves 9.8% improvement in long-term memory retention. |
Efficient Multilingual Multi-modal Pre-training through Triple Contrastive Loss (2022.coling-1)
Copied to clipboard
| Challenge: | Existing approaches to learn visual and textual representations from web-scale image-text pairs are limited due to labeling cost and limited scalability. |
| Approach: | They propose to use web-scale image-text pairs to learn visual and textual representations in the shared space. |
| Outcome: | The proposed enhancement scheme improves multilingual vision-and-language tasks by minimizing a triplet contrastive loss on images and two different language texts with the same meaning. |
Can LLMs Truly Plan? A Comprehensive Evaluation of Planning Capabilities (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing assessments of planning capabilities of large language models are limited to single-language or specific representation formats. |
| Approach: | a new benchmark is developed to assess the planning capabilities of large language models. |
| Outcome: | The Multi-Plan benchmark highlights performance disparities among models . language differences showed minimal impact, while mathematically structured representations improved accuracy . |
ScholarBench: A Bilingual Benchmark for Abstraction, Comprehension, and Reasoning Evaluation in Academic Contexts (2025.findings-emnlp)
Copied to clipboard
| Challenge: | ScholarBench evaluates domain-specific knowledge of large language models (LLMs) prior benchmarks lack the scalability to handle complex academic tasks. |
| Approach: | ScholarBench evaluates the academic reasoning ability of large language models . the benchmark is constructed through a three-step process . |
| Outcome: | ScholarBench evaluates the academic reasoning ability of large language models . the benchmark comprises 5,031 examples in Korean and 5,309 examples in English . |
Semantic Hardness Is Not Visual Hardness: Sign-Aware Hard Negative Mining for Sign Language Retrieval (2026.acl-long)
Copied to clipboard
Junmyeong Lee, Chan Hur, ChangSu Choi, Sukmin Cho, Fitsum Gaim, Eui Jun Hwang, Hoyun Song, KyungTae Lim
| Challenge: | Existing methods for sign language retrieval fail to capture visual ambiguity . semantically distinct yet visually confusable signs are rarely treated as hard negatives . |
| Approach: | They propose a method that constructs hard negatives based on visual confusability rather than linguistic similarity. |
| Outcome: | The proposed method significantly improves fine-grained retrieval performance while preserving coarse-grain accuracy. |
VLR-Bench: Multilingual Benchmark Dataset for Vision-Language Retrieval Augmented Generation (2025.coling-main)
Copied to clipboard
Hyeonseok Lim, Dongjae Shin, Seohyun Song, Inho Won, Minjun Kim, Junghun Yuk, Haneol Jang, KyungTae Lim
| Challenge: | Existing evaluation datasets for external knowledge-based VQA lack a capability to determine which passage is useful for answering queries. |
| Approach: | They propose a visual question answering benchmark for vision language models based on retrieval augmented generation (RAG) the proposed benchmark includes five input passages, a capability lacking in previous research. |
| Outcome: | The proposed benchmark includes five input passages and is validated using the state-of-the-art Llama3-based VLM, the Llava-Llamama-3 model. |
Multilingual Dependency Parsing for Low-Resource Languages: Case Studies on North Saami and Komi-Zyrian (L18-1)
Copied to clipboard
| Challenge: | Developing systems for low-resource languages is a crucial issue for Natural Language Processing (NLP). |
| Approach: | They propose a method for parsing low-resource languages with very small training corpora using multilingual word embeddings and annotated corporata of larger languages. |
| Outcome: | The proposed method improves dependency parsing for low-resource languages with very small training corpora compared to previous work . it also explores whether contemporary contact languages or genetically related languages would be the most fruitful starting point for multilingual parsers. |
Towards Standardized Annotation and Parsing for Korean FrameNet (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing studies on Korean FrameNet have focused on English, but annotations are not optimally designed for Korean. |
| Approach: | They propose a morphologically enhanced annotation strategy for Korean FrameNet datasets and parsing by leveraging the CoNLL-U format. |
| Outcome: | The proposed method improves the annotation accuracy of Korean FrameNet datasets and their parsers. |
When the Misidentified Adverbial Phrase Functions as a Complement (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Existing studies on the predicate-argument structure of natural languages have been limited . despite the importance of distinguishing mandatory arguments and optional modifiers, there is no dataset with token-level annotations for Korean . |
| Approach: | They propose a dataset which labels mandatory and optional elements as complements and adjuncts in Korean . they reclassify Korean phrases as complement and develop an automatic labeling technique for complements . |
| Outcome: | The proposed dataset is trainable and reliable, with satisfying results . Korean is an adverbial language, and its predicate-argument structure is well understood . |
Optimizing Language Augmentation for Multilingual Large Language Models: A Case Study on Korean (2024.lrec-main)
Copied to clipboard
ChangSu Choi, Yongbin Jeong, Seoyoon Park, Inho Won, HyeonSeok Lim, SangMin Kim, Yejee Kang, Chanhyuk Yoon, Jaewan Park, Yiseul Lee, HyeJin Lee, Younggyun Hahm, Hansaem Kim, KyungTae Lim
| Challenge: | Large language models (LLMs) use pretraining to predict the subsequent word, but less-resourced languages are being overlooked. |
| Approach: | They propose to expand the MLLM vocabularies to enhance expressiveness and use bilingual data for pretraining to align the high- and less-resourced languages. |
| Outcome: | The proposed model outperforms existing models in qualitative analyses compared to Korean monolingual models. |
ELO: Efficient Layer-Specific Optimization for Continual Pretraining of Multilingual LLMs (2026.eacl-industry)
Copied to clipboard
Hangyeol Yoo, ChangSu Choi, Minjun Kim, Seohyun Song, SeungWoo Song, Inho Won, Jongyoul Park, Cheoneum Park, KyungTae Lim
| Challenge: | Recent studies have focused on enhancing multilingual large language models (MLLMs) for specific languages. |
| Approach: | They propose an efficient layer-specific optimization method to enhance continual pretraining (CP) for specific languages in multilingual large language models (MLLMs). |
| Outcome: | The proposed method achieves a training speedup of up to 6.46 times compared to existing methods while improving target language performance by up to 5.2% on qualitative benchmarks. |
SCV: Light and Effective Multi-Vector Retrieval with Sequence Compressive Vectors (2025.coling-industry)
Copied to clipboard
| Challenge: | Recent advances in language models (LMs) have driven progress in information retrieval (IR), effectively extracting semantically relevant information. |
| Approach: | They propose a multi-vector retrieval framework with sequence compression vectors and a coarse-to-fine vector search framework that uses span compressive vectors to score documents. |
| Outcome: | The proposed approach achieves the fastest latency compared to other state-of-the-art models and can obtain competitive performance on both in-domain and out-of domain benchmark datasets. |
Unified Automated Essay Scoring and Grammatical Error Correction (2025.findings-naacl)
Copied to clipboard
| Challenge: | a new study explores the integration of automated writing evaluation and grammatical error correction through multitask learning. |
| Approach: | They propose a system that integrates automated writing evaluation and grammatical error correction through multitask learning by leveraging a shared learning framework. |
| Outcome: | The proposed system outperforms models trained on AWE and GEC, the authors show . their study demonstrates that the proposed system improves writing assessment accuracy and accuracy . |