5 citations · 7 across the 18 of their papers we have counts for
19 papers
Small Models Scout Bottleneck Order for Large-Model Data Control
Seungmin Choi, Jiwon Sung, Muhammad Umer +4
Small proxy models are commonly used to identify data mixtures for larger-scale training. We ask whether their training trajectories reveal another transferable structure: the orde…
Self-Improving CAD Generation Agents with Finite Element Analysis as Feedback
Guijin Son, Jehyun Park, Seyeon Park +2
Computer-aided design (CAD) is the backbone of modern industrial design, yet learned CAD generators still fall short of real engineering pipelines: they neither iterate like engine…
Soohak: A Mathematician-Curated Benchmark for Evaluating Research-level Math Capabilities of LLMs
Guijin Son, Seungone Kim, Catherine Arnett +73
Following the recent achievement of gold-medal performance on the IMO by frontier LLMs, the community is searching for the next meaningful and challenging target for measuring LLM…
KMMMU: Evaluation of Massive Multi-discipline Multimodal Understanding in Korean Language and Context
Nahyun Lee, Guijin Son, Hyunwoo Ko +4
Multimodal evaluation is strengthened when benchmarks reflect the languages, institutions, and disciplinary conventions that shape expert visual reasoning. We introduce KMMMU, a na…
Pushing the Boundaries of Multiple Choice Evaluation to One Hundred Options
Nahyun Lee, Guijin Son
Multiple choice evaluation is widely used for benchmarking large language models, yet near ceiling accuracy in low option settings can be sustained by shortcut strategies that obsc…
Judging What We Cannot Solve: A Consequence-Based Approach for Oracle-Free Evaluation of Research-Level Math
Guijin Son, Donghun Yang, Hitesh Laxmichand Patel +5
Recent progress in reasoning models suggests that generating plausible attempts for research-level mathematics may be within reach, but verification remains a bottleneck, consuming…
What Users Leave Unsaid: Under-Specified Queries Limit Vision-Language Models
Dasol Choi, Guijin Son, Hanwool Lee +7
Current vision-language benchmarks predominantly feature well-structured questions with clear, explicit prompts. However, real user queries are often informal and underspecified. U…
Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought
Guijin Son, Donghun Yang, Hitesh Laxmichand Patel +9
Recent frontier models employ long chain-of-thought reasoning to explore solution spaces in context and achieve stonger performance. While many works study distillation to build sm…
Ko-PIQA: A Korean Physical Commonsense Reasoning Dataset with Cultural Context
Dasol Choi, Jungwhan Kim, Guijin Son
Physical commonsense reasoning datasets like PIQA are predominantly English-centric and lack cultural diversity. We introduce Ko-PIQA, a Korean physical commonsense reasoning datas…
KAIO: A Collection of More Challenging Korean Questions
Nahyun Lee, Guijin Son, Hyunwoo Ko +1
With the advancement of mid/post-training techniques, LLMs are pushing their boundaries at an accelerated pace. Legacy benchmarks saturate quickly (e.g., broad suites like MMLU ove…
When AI Co-Scientists Fail: SPOT-a Benchmark for Automated Verification of Scientific Research
Guijin Son, Jiwoo Hong, Honglu Fan +8
Recent advances in large language models (LLMs) have fueled the vision of automated scientific discovery, often called AI Co-Scientists. To date, prior work casts these systems as…
On the Robustness of Reward Models for Language Model Alignment
Jiwoo Hong, Noah Lee, Eunki Kim +5
The Bradley-Terry (BT) model is widely practiced in reward modeling for reinforcement learning with human feedback (RLHF). Despite its effectiveness, reward models (RMs) trained wi…
Controlling Language Confusion in Multilingual LLMs
Nahyun Lee, Yeongseo Woo, Hyunwoo Ko +1
Large language models often suffer from language confusion, a phenomenon in which responses are partially or entirely generated in unintended languages. This critically degrades th…
Won: Establishing Best Practices for Korean Financial NLP
Guijin Son, Hyunwoo Ko, Haneral Jung +1
In this work, we present the first open leaderboard for evaluating Korean large language models focused on finance. Operated for about eight weeks, the leaderboard evaluated 1,119…
Linguistic Generalizability of Test-Time Scaling in Mathematical Reasoning
Guijin Son, Jiwoo Hong, Hyunwoo Ko +1
Scaling pre-training compute has proven effective for achieving mulitlinguality, but does the same hold for test-time scaling? In this work, we introduce MCLM, a multilingual math…
Understand, Solve and Translate: Bridging the Multilingual Mathematical Reasoning Gap
Hyunwoo Ko, Guijin Son, Dasol Choi
Large language models (LLMs) demonstrate exceptional performance on complex reasoning tasks. However, despite their strong reasoning capabilities in high-resource languages (e.g.,…
Multi-Step Reasoning in Korean and the Emergent Mirage
Guijin Son, Hyunwoo Ko, Dasol Choi
We introduce HRMCR (HAE-RAE Multi-Step Commonsense Reasoning), a benchmark designed to evaluate large language models' ability to perform multi-step reasoning in culturally specifi…
Improving Fine-grained Visual Understanding in VLMs through Text-Only Training
Dasol Choi, Guijin Son, Soo Yong Kim +2
Visual-Language Models (VLMs) have become a powerful tool for bridging the gap between visual and linguistic understanding. However, the conventional learning approaches for VLMs o…
MM-Eval: A Multilingual Meta-Evaluation Benchmark for LLM-as-a-Judge and Reward Models
Guijin Son, Dongkeun Yoon, Juyoung Suk +7
As Large Language Models (LLMs) are now capable of producing fluent and coherent content in languages other than English, it is not imperative to precisely evaluate these non-Engli…