Papers by Hwanjo Yu
OoMMix: Out-of-manifold Regularization in Contextual Embedding Space for Text Classification (2021.acl-long)
Copied to clipboard
| Challenge: | Recent studies on neural networks with pre-trained weights focus on low-dimensional subspace where the embedding vectors computed from input words are located. |
| Approach: | They propose an approach to find and regularize the remainder of the space, referred to as out-of-manifold, which cannot be accessed through the words. |
| Outcome: | The proposed approach is able to fine-tune the out-of-manifold embedding space on text classification benchmarks. |
Out-of-domain Detection based on Generative Adversarial Network (D18-1)
Copied to clipboard
| Challenge: | Existing methods for out-of-domain (OOD) detection require huge effort to collect OOD sentences. |
| Approach: | They propose to use only in-domain (IND) sentences to build a generative adversarial network (GAN) of which the discriminator generates low scores for OOD sentences. |
| Outcome: | The proposed method is most accurate compared to existing methods on multi-domain dialog systems. |
Building Large-Scale English and Korean Datasets for Aspect-Level Sentiment Analysis in Automotive Domain (2020.coling-main)
Copied to clipboard
| Challenge: | Existing datasets in automotive domain cover only three languages due to high cost of human annotation. |
| Approach: | They build large-scale datasets of users’ comments in two languages, English and Korean, for aspect-level sentiment analysis in automotive domain. |
| Outcome: | The datasets consist of 58,000+ commentaspect pairs, which are the largest compared to existing datasets. |
Rectifying Demonstration Shortcut in In-Context Learning (2024.naacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) can solve tasks with a few demonstrations, but often rely on their pre-trained semantic priors rather than the input-label relationships to proceed with ICL prediction. |
| Approach: | They propose a demonstration-aware calibration method to improve LLMs' ability to learn new input-label relationships from demonstrations. |
| Outcome: | The proposed method improves the original ICL task and the task learning setting, and the results are generalized across three LLM families. |
Verbosity-Aware Rationale Reduction: Sentence-Level Rationale Reduction for Efficient and Effective Reasoning (2025.findings-acl)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) demonstrate remarkable reasoning capabilities comparable to human cognitive abilities. |
| Approach: | They propose a sentence-level rationale reduction framework that leverages likelihood-based criteria to selectively remove redundant reasoning sentences while preserving reasoning capabilities. |
| Outcome: | The proposed framework improves performance by 7.71% while reducing token generation by 19.87% compared to models trained with complete reasoning paths. |
Exploring Language Model’s Code Generation Ability with Auxiliary Functions (2024.findings-naacl)
Copied to clipboard
| Challenge: | Auxiliary function is a useful component to improve language model’s code generation ability, but a systematic exploration of how they affect has yet to be done. |
| Approach: | They construct a human-crafted evaluation set which contains examples of two functions where one function assists the other to examine their ability in a multifaceted way. |
| Outcome: | The proposed model is underutilized to call the auxiliary function, suggesting future directions to enhance their implementation by eliciting the supplementary function call ability encoded in the models. |
Valid Necessary: Diagnosing Latent Inefficiency in Chain-of-Thought (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing reasoning step evaluators fail to distinguish “valid but inefficient” reasoning steps from necessary reasoning. |
| Approach: | They propose a training-free metric that identifies low-utility steps and a post-hoc compression strategy to quantify their impact on token usage. |
| Outcome: | The proposed metric reduces token consumption by 31–53% while maintaining accuracy at substantially higher compression rates. |
Taxonomy-guided Semantic Indexing for Academic Paper Search (2024.emnlp-main)
Copied to clipboard
| Challenge: | Academic paper search often struggles to match underlying academic concepts between queries and documents. |
| Approach: | They propose a framework that extracts key concepts from papers and organizes them as a semantic index guided by an academic taxonomy. |
| Outcome: | The proposed framework can be flexibly employed to enhance existing retrieval frameworks. |
Are Vision-Language Models Safe in the Wild? A Meme-Based Benchmark Study (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing safety evaluations rely on artificial images to evaluate vision-language models . a recent study found that memes are more effective at bypassing safety measures than synthetic or typographic images. |
| Approach: | They propose a benchmark pairing meme images with harmful and benign instructions . they assess multiple VLMs across single and multi-turn interactions . |
| Outcome: | The proposed benchmark pairs real meme images with harmful and benign instructions. |
COMPASS: A Framework for Evaluating Organization-Specific Policy Alignment in LLMs (2026.acl-long)
Copied to clipboard
Dasol Choi, DongGeon Lee, Brigitta Jesica Kartono, Helena Berndt, Taeyoun Kwon, Joonwon Jang, Haon Park, Hwanjo Yu, Minsuk Kahng
| Challenge: | Large language models are being rapidly adopted across a wide range of domains, including healthcare, finance, and the public sector. |
| Approach: | They propose a framework to evaluate whether large language models comply with policies . they apply COMPASS to eight diverse industry scenarios to validate models . |
| Outcome: | The proposed framework evaluates whether LLMs comply with allowlist and denylist policies. |
Toward Interpretable Semantic Textual Similarity via Optimal Transport-based Contrastive Sentence Learning (2022.acl-long)
Copied to clipboard
| Challenge: | Existing studies have shown that a pretrained language model can capture sentence similarity but there is no interpretation method for the sentence similarities. |
| Approach: | They propose a pretrained language model that captures sentence similarity between embeddings and a transport-based distance measure that leverages semantically-aligned token pairs. |
| Outcome: | The proposed framework outperforms baselines on both STS and interpretable-STS benchmarks and provides interpretation consistent with human judgement. |
Exploring Iterative Controllable Summarization with Large Language Models (2026.findings-eacl)
Copied to clipboard
| Challenge: | Large language models (LLMs) excel at abstractive summarization tasks, but their ability to precisely control summary attributes remains underexplored. |
| Approach: | They propose a guide-to-explain framework for controllable summarization that enables the model to identify misaligned attributes in the initial draft and guides it to self-explan errors in the previous output. |
| Outcome: | The proposed framework generates well-adjusted summaries that satisfy the desired attributes with robust effectiveness while requiring surprisingly fewer iterations than other iterative approaches. |
Eliciting Instruction-tuned Code Language Models’ Capabilities to Utilize Auxiliary Function for Code Generation (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Using auxiliary functions to implement functions is important for instruction-tuned models because it reduces the implementation difficulty of a target function compared to implementing them from scratch. |
| Approach: | They propose several ways to provide auxiliary functions to the models by adding them to the query or providing a response prefix to incorporate the ability to utilize auxiliary function with the instruction following capability. |
| Outcome: | The proposed models outperform the recent powerful language models, gpt-4o, in the code generation task. |
KoDialogBench: Evaluating Conversational Understanding of Language Models with Korean Dialogue Benchmark (2024.lrec-main)
Copied to clipboard
| Challenge: | KoDialogBench is a benchmark designed to assess language models’ conversational capabilities in low-resource languages such as Korean. |
| Approach: | They propose a benchmark to assess language models’ conversational capabilities in Korean by collecting native Korean dialogues from public sources and translating them into diverse test datasets. |
| Outcome: | The proposed benchmark measures the conversational capabilities of language models in Korean, and shows that they can improve on previous training techniques. |
Generating Multiple-Length Summaries via Reinforcement Learning for Unsupervised Sentence Summarization (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing models to summarize texts without ground-truth summaries are extractive, which remove words from texts and thus are less flexible than abstractive models. |
| Approach: | They propose an unsupervised model that extracts words from texts and makes them mutually enhance each other. |
| Outcome: | The proposed model outperforms both abstractive and extractive models, while generating new words not contained in input texts. |
Topic Taxonomy Expansion via Hierarchy-Aware Topic Phrase Generation (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods for topic taxonomies focus on frequent terms and local topic-subtopic relations, which leads to limited topic term coverage. |
| Approach: | They propose a framework for topic taxonomy expansion that directly generates topic-related terms belonging to new topics. |
| Outcome: | The proposed framework outperforms baseline methods on two real-world text corpora. |
Topic Coverage-based Demonstration Retrieval for In-Context Learning (2025.emnlp-main)
Copied to clipboard
| Challenge: | Prior methods to retrieve demonstrations based on embedding similarity or generation probability, resulting in irrelevant or redundant examples. |
| Approach: | They propose a topic coverage-based retrieval framework that selects demonstrations to comprehensively cover topic-level knowledge relevant to both the test input and the model. |
| Outcome: | The proposed framework covers all the necessary knowledge for the test input and the model. |
StepER: Step-wise Knowledge Distillation for Enhancing Reasoning Ability in Multi-Step Retrieval-Augmented Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing knowledge distillation methods overlook the need for different reasoning abilities at different steps, hindering transfer in multi-step retrieval-augmented frameworks. |
| Approach: | They propose a method that uses step-wise supervision to align with evolving information and reasoning demands across stages. |
| Outcome: | The proposed method outperforms previous methods on multi-hop QA benchmarks with an 8B model achieving performance comparable to a 70B teacher model. |
How Diversely Can Language Models Solve Problems? Exploring the Algorithmic Diversity of Model-Generated Code (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Language models (LMs) have exhibited impressive abilities in generating code from natural language requirements. |
| Approach: | They propose to introduce various metrics with inter-code similarity to evaluate the diversity of generated code by comparing model-generated solutions with human-written ones. |
| Outcome: | The proposed method leverages LMs’ capabilities in code understanding and reasoning, resulting in a set of metrics that represent the number of algorithms in model-generated solutions. |