Papers by Kang Kim
Restoring and Mining the Records of the Joseon Dynasty via Neural Language Modeling and Machine Translation (2021.naacl-main)
Copied to clipboard
| Challenge: | voluminous historical records are difficult to fully utilize since they are written in ancient languages and some parts are damaged over time. |
| Approach: | They propose a multi-task learning approach to restore and translate historical documents using a self-attention mechanism. |
| Outcome: | The proposed approach improves the accuracy of the translation task over baselines without multi-task learning. |
SLM as Guardian: Pioneering AI Safety with Small Language Model (2024.emnlp-industry)
Copied to clipboard
Ohjoon Kwon, Donghyeon Jeon, Nayoung Choi, Gyu-Hwung Cho, Hwiyeol Jo, Changbong Kim, Hyunwoo Lee, Inho Kang, Sun Kim, Taiwoo Park
| Challenge: | Prior safety research on large language models focused on aligning them to safety requirements, but internalizing such safeguard features into larger models brought challenges of higher training cost and unintended degradation of helpfulness. |
| Approach: | They propose a multi-task learning mechanism that integrates harmful query detection and safeguard response into a single model. |
| Outcome: | The proposed approach outperforms the publicly available LLMs in harmful query detection and safeguard response generation. |
Generating Information-Seeking Conversations from Unlabeled Documents (2022.emnlp-main)
Copied to clipboard
| Challenge: | a novel framework for conversational question answering from unlabeled documents has been proposed . a large-scale dataset of synthetic conversations is available for use in real-world applications . |
| Approach: | They propose a framework for conversational question answering from unlabeled documents . they propose 'SimSeek' framework that simulates conversation from unlabelled documents based on two scenarios . |
| Outcome: | The proposed framework achieves state-of-the-art performance on a recent CQA benchmark, QuAC. |
Rationale-Guided Retrieval Augmented Generation for Medical Question Answering (2025.naacl-long)
Copied to clipboard
Jiwoong Sohn, Yein Park, Chanwoong Yoon, Sihyeon Park, Hyeon Hwang, Mujeen Sung, Hyunjae Kim, Jaewoo Kang
| Challenge: | Large language models (LLMs) struggle with hallucinations and outdated knowledge. |
| Approach: | They propose a retrieval-augmented generation framework for enhancing the reliability of RAG in biomedical contexts. |
| Outcome: | The proposed framework outperforms the previous best medical RAG model by up to 5.6% across three medical question-answering benchmarks. |
Human-in-the-Loop Synthetic Text Data Inspection with Provenance Tracking (2024.findings-naacl)
Copied to clipboard
| Challenge: | Data augmentation techniques generate low-quality texts with incorrect labels . a new technique is needed to winnow out texts with inaccurate labels based on provenance inspection . |
| Approach: | They develop a data inspection technique that uses provenance inspection and assistive labeling to winnow out texts with incorrect labels. |
| Outcome: | a new human-in-the-loop data inspection technique can winnow out texts with incorrect labels . the technique can reduce human inspection effort by combining provenance inspection and assistive labeling . |
“Why do I feel offended?” - Korean Dataset for Offensive Language Identification (2023.findings-eacl)
Copied to clipboard
| Challenge: | Existing methods for detecting offensive content rely on labeled datasets, but few consider low-resource languages with relatively less data available for training. |
| Approach: | They propose to use Korean as a dataset for offensive language identification . they propose to perform abusive language detection and sentiment analysis to help identify offensive languages. |
| Outcome: | The proposed datasets improve the performance of offensive language identification in Korean, while the existing methods are limited. |
FQ-Eval: Building Evaluation Dataset for User-centered Follow-up Question Generation (2025.emnlp-industry)
Copied to clipboard
| Challenge: | Existing studies focus on enhancing information-seeking or topical relevance, often missing how follow-up questions could satisfy users’ intrinsic needs and conversational goals. |
| Approach: | They propose a user-centered evaluation dataset for assessing follow-up question generation in chat-LLM services that incorporates realistic chat-llm usage scenarios and five distinct human-aligned criteria. |
| Outcome: | The proposed model captures human-aligned criteria for the evaluation of various models and services. |
Automatic Creation of Named Entity Recognition Datasets by Querying Phrase Representations (2023.acl-long)
Copied to clipboard
| Challenge: | Named entity recognition models rely on domain-specific dictionaries provided by experts . however, such dictionary sets are infeasible in many domains where they do not exist . |
| Approach: | They propose a framework that generates NER datasets with high-coverage pseudo-dictionaries . phrase retrieval models are used to retrieve popular entities rather than rare ones . |
| Outcome: | The proposed framework outperforms the previous best model by an average F1 score of 4.7 across five NER benchmark datasets. |
Korean Language Modeling via Syntactic Guide (2022.lrec-1)
Copied to clipboard
| Challenge: | Existing research on pre-trained language models focuses on widely-used languages . however, not every language can benefit from such models due to computational resources . |
| Approach: | They propose to build a pre-trained language model that understands the linguistic phenomena in the target language with low resources. |
| Outcome: | The proposed model improves the performance of Korean language understanding tasks. |
Benchmarking Direct Preference Optimization for Medical Large Vision–Language Models (2026.findings-eacl)
Copied to clipboard
| Challenge: | Large vision-language models (LVLMs) are gaining traction in clinical tasks such as diagnostic support, report generation, and medical question answering. |
| Approach: | They present a systematic evaluation of nine DPO variants applied to two leading medical LVLMs. |
| Outcome: | The proposed model improves alignment and reduces severe hallucinations, but yields inconsistent gains over supervised fine-tuning. |
Riemannian Optimization for LoRA on the Stiefel Manifold (2025.findings-emnlp)
Copied to clipboard
| Challenge: | powerful, large language models (LLMs) present significant fine-tuning challenges due to their size. |
| Approach: | They propose to optimize LoRA’s B matrix by imposing explicit orthogonality constraints that achieve near-perfect orthogonal and full effective rank. |
| Outcome: | The proposed method outperforms AdamW and LoRA in terms of parameter efficiency and representational capacity. |
ETHIC: Evaluating Large Language Models on Long-Context Tasks with High Information Coverage (2025.naacl-long)
Copied to clipboard
| Challenge: | Existing evaluation methods do not assess whether large language models fully utilize contextual information. |
| Approach: | They introduce a new metric to assess LLMs' ability to fully utilize contextual information. |
| Outcome: | The proposed benchmark comprises 1,986 test instances spanning four long-context tasks with high IC scores in the domains of books, debates, medicine, and law. |
Ranking Paragraphs for Improving Answer Recall in Open-Domain Question Answering (D18-1)
Copied to clipboard
| Challenge: | Recent work has combined open-domain question answering with machine comprehension models to find answers in a large knowledge source. |
| Approach: | They propose a machine comprehension model that ranks paragraphs of retrieved documents for a higher answer recall with less noise. |
| Outcome: | The proposed model improves on four open-domain QA datasets by 7.8% on average. |
CPC-GRPO: Answer-Free Reinforcement Learning with Cross-Prompt Consensus Rewards (2026.findings-acl)
Copied to clipboard
| Challenge: | Reinforcement learning with verifiable rewards is a popular post-training tool for large language models, but relies on a ground-truth answer or external verifier, which limits applicability and increases cost. |
| Approach: | They propose an answer-free training objective that derives rewards solely from the model’s own probabilities by exploiting prompt paraphrases as multiple semantic views of the same intent. |
| Outcome: | The proposed objective derives rewards solely from the model’s own probabilities by exploiting prompt paraphrases as multiple semantic views of the same intent. |
Thunder-NUBench: A Benchmark for LLMs’ Sentence-Level Negation Understanding (2026.findings-eacl)
Copied to clipboard
| Challenge: | Negation is a fundamental linguistic phenomenon that poses ongoing challenges for Large Language Models (LLMs) Current benchmarks treat negation as a minor detail within broader tasks, such as natural language inference. |
| Approach: | They propose a novel benchmark specifically created to assess sentence-level understanding of negation in Large Language Models (LLMs). |
| Outcome: | The proposed benchmark compares standard negation with structurally diverse alternatives, such as local negation, contradiction, and paraphrase. |
MM-PoisonRAG: Disrupting Multimodal RAG with Local and Global Knowledge Poisoning Attacks (2026.acl-long)
Copied to clipboard
Hyeonjeong Ha, Qiusi Zhan, Jeonghwan Kim, Dimitrios Bralios, Saikrishna Sanniboina, Nanyun Peng, Kai-Wei Chang, Daniel Kang, Heng Ji
| Challenge: | Existing research exposes multimodal large language models to knowledge poisoning attacks . localized poisoning attack achieves up to 56% success rate even under restricted access . globalized poison attack completely disrupts model generation to 0% accuracy with just one poisoned content. |
| Approach: | They propose a framework to study the vulnerability of multimodal RAG under knowledge poisoning attacks. |
| Outcome: | The proposed framework exploits two new attack strategies on multimodal RAGs under knowledge poisoning. |
MATE: Meet At The Embedding - Connecting Images with Long Texts (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Recent advances in Vision Language Models (VLMs) focus on aligning images with short descriptive captions. |
| Approach: | They propose a method that combines VLMs with Large Language Models to efficiently align images with long texts without additional text pairs. |
| Outcome: | The proposed method bridges the gap between VLM and LLM without additional image-long text pairs. |
Exploiting Numerical-Contextual Knowledge to Improve Numerical Reasoning in Question Answering (2022.findings-naacl)
Copied to clipboard
| Challenge: | Existing numerical reasoning models overly rely on parametric knowledge at inference time . previous studies show that understanding numbers in text improves numerical reasoning accuracy . |
| Approach: | They propose a numerical reasoning model that leverages parametric knowledge to alleviate this over-reliance on parametric information. |
| Outcome: | The proposed model improves numerical reasoning accuracy and performance in DROP. |
One Missing Piece for Open-Source Reasoning Models: A Dataset to Mitigate Cold-Starting Short CoT LLMs in RL (2025.acl-industry)
Copied to clipboard
Hyungjoo Chae, Dongjin Kang, Jihyuk Kim, Beong-woo Kwak, Sunghyun Park, Haeju Park, Jinyoung Yeo, Moontae Lee, Kyungjae Lee
| Challenge: | Existing large reasoning models are limited by their closed nature and high API costs and safety issues. |
| Approach: | They propose to build a long CoT dataset with existing short CoT LLMs that are not trained for inference-time scaling. |
| Outcome: | The proposed model achieves quality comparable to—or slightly below—R1 and is able to think longer and provide control over the thought budget to better manage the overthinking problem. |
Generation-Based and Emotion-Reflected Memory Update: Creating the KEEM Dataset for Better Long-Term Conversation (2025.coling-main)
Copied to clipboard
| Challenge: | KEEM is a dynamically generated dataset designed to enhance memory updates in long-term conversational systems. |
| Approach: | They propose a dataset that keeps emotional and essential memories and generates integrative memories that incorporate emotional context and causal relationships. |
| Outcome: | The Keep Emotional and Essential Memory (KEEM) dataset enhances memory updates in long-term conversational systems. |
SWITCH: Studying with Teacher for Knowledge Distillation of Large Language Models (2025.findings-naacl)
Copied to clipboard
| Challenge: | Knowledge Distillation (KD) has emerged as a popular method for compressing large language models due to high inference costs and memory requirements. |
| Approach: | They propose a method that integrates the teacher model during the student's sequence generation to reduce misguidance from the teacher. |
| Outcome: | Experiments on three model families and five instruction-following datasets show that SWITCH surpasses traditional methods, especially in the generation of long sequential data. |
Revisiting the Impact of Pursuing Modularity for Code Generation (2024.findings-emnlp)
Copied to clipboard
| Challenge: | a recent study examines the impact of modularity on code generation in large language models . modularity is not a core factor for improving performance of code generation models, argues a new study . |
| Approach: | They introduce a new metric to measure the impact of modularity in code generation . they find modularity is not a core factor for improving performance of LLMs . |
| Outcome: | The proposed metric shows that modularity is not a core factor for improving performance . coding assistants are becoming increasingly essential for programmers . |
Pearl: A Review-driven Persona-Knowledge Grounded Conversational Recommendation Dataset (2024.findings-acl)
Copied to clipboard
Minjin Kim, Minju Kim, Hana Kim, Beong-woo Kwak, SeongKu Kang, Youngjae Yu, Jinyoung Yeo, Dongha Lee
| Challenge: | Existing datasets for conversational recommender systems lack specific user preferences and explanations for recommendations . current datasets lack specific preferences, hindering high-quality recommendations despite advances in large language models . |
| Approach: | They propose to synthesize a conversational recommendation dataset with persona- and knowledge-augmented LLM simulators to address these challenges. |
| Outcome: | The proposed dataset outperforms baselines in human and automatic evaluations. |
TRIAL: Token Relations and Importance Aware Late-interaction for Accurate Text Retrieval (2025.emnlp-main)
Copied to clipboard
| Challenge: | Late-interaction based multi-vector retrieval systems rely on a naive summation of token-level similarity scores . this leads to inaccurate relevance estimation due to tokenization of semantic units and the influence of low-content words. |
| Approach: | They propose a late-interaction-based multi-vector retrieval system that uses token relations and token importance in relevance scoring. |
| Outcome: | Extensive tests show that TRIAL achieves state-of-the-art accuracy compared to existing methods. |
Why Do Multilingual Reasoning Gaps Emerge in Reasoning Language Models? (2026.findings-acl)
Copied to clipboard
| Challenge: | Recent reasoning language models (RLMs) achieve strong performance on complex reasoning tasks, yet they still exhibit a multilingual reasoning gap. |
| Approach: | They propose a strategy that incorporates an English translation into the initial reasoning trace when an understanding failure is detected. |
| Outcome: | The proposed strategy incorporates an English translation into the initial reasoning trace when an understanding failure is detected. |
Sightation Counts: Leveraging Sighted User Feedback in Building a BLV-aligned Dataset of Diagram Descriptions (2025.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that direct generation of diagram descriptions is costly and biased against blind and low-vision (BLV) users. |
| Approach: | They ask sighted individuals to assess diagram descriptions generated by vision-language models . they use latent supervision to guide the models with latent inference . |
| Outcome: | The results show that visual descriptions generated by vision-language models are effective and useful to educators who are themselves BLV and teach visually impaired learners. |
Don’t Judge Code by Its Cover: Exploring Biases in LLM Judges for Code Evaluation (2026.findings-eacl)
Copied to clipboard
| Challenge: | Large language models (LLMs) are increasingly used as evaluators for code evaluation tasks . however, whether they can handle superficial variations remains unclear . |
| Approach: | They define six types of potential biases in code evaluation and reveal their impact on LLM judges. |
| Outcome: | The proposed method can be used to evaluate semantically equivalent code with superficial variations without reference implementations. |
Learn to Resolve Conversational Dependency: A Consistency Training Framework for Conversational Question Answering (2021.acl-long)
Copied to clipboard
| Challenge: | Existing approaches do not explicitly train QA models on how to resolve conversational dependency, and thus these models are limited in understanding human dialogues. |
| Approach: | They propose a framework that generates self-contained questions that can be understood without the conversation history and then trains a QA model with the pairs of original and self-constructed questions using a consistency-based regularizer. |
| Outcome: | The proposed framework improves the models’ performance by up to 1.2 F1 on QuAC, and 5.2 F1 for CANARD, while addressing the limitations of the existing approaches. |
Have You Seen That Number? Investigating Extrapolation in Question Answering Models (2021.emnlp-main)
Copied to clipboard
| Challenge: | Numerical reasoning in machine reading comprehension (MRC) has shown drastic improvements over the past few years. |
| Approach: | They propose an E-digit number form that alleviates the lack of extrapolation in numerical MRC models. |
| Outcome: | The proposed model can't extrapolate to unseen numbers, the authors say . they also show that the model needs to treat numbers differently from regular words . |
Extending CLIP’s Image-Text Alignment to Referring Image Segmentation (2024.naacl-long)
Copied to clipboard
| Challenge: | Referring Image Segmentation (RIS) is a cross-modal task that aims to segment an instance described by a natural language expression. |
| Approach: | They propose a framework that leverages the cross-modal nature of CLIP for RIS by leveraging image-text alignment knowledge in CLIP's image-embedding space. |
| Outcome: | The proposed framework outperforms CLIP-based methods on all three major RIS benchmarks and outperformed previous CLIP methods. |
LM-BFF-MS: Improving Few-Shot Fine-tuning of Language Models based on Multiple Soft Demonstration Memory (2022.acl-short)
Copied to clipboard
| Challenge: | LM-BFF (CITATION) achieves significant few-shot performance by using auto-generated prompts and adding demonstrations similar to an input example. |
| Approach: | They propose to use auto-generated prompts and add demonstrations to LM-BFF to improve few-shot fine-tuning of language models with multiple soft demonstrations. |
| Outcome: | The proposed method improves few-shot fine-tuning on eight NLP tasks. |
Fooling the LVLM Judges: Visual Biases in LVLM-Based Evaluation (2025.emnlp-main)
Copied to clipboard
| Challenge: | a recent study has found that large vision–language models are vulnerable to visual biases that inflate scores without altering semantic content. |
| Approach: | They propose a novel meta-evaluation benchmark that exhibits diverse score distributions. |
| Outcome: | The proposed model exhibits vulnerability across all domains, and combines multiple biases amplifies their effects, and pairwise evaluations are similarly susceptible. |
Large Language Models are Temporal and Causal Reasoners for Video Question Answering (2023.emnlp-main)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown remarkable performances on a wide range of natural language understanding and generation tasks. |
| Approach: | They propose a framework that exploits linguistic shortcuts and mitigates 'linguistic bias' by flipping the source pair and target label to understand their complex relationships. |
| Outcome: | The proposed framework outperforms both LLMs-based and non-LLMs- based models on five challenging VideoQA benchmarks. |
Self-Correcting Code Generation Using Small Language Models (2025.findings-emnlp)
Copied to clipboard
| Challenge: | a recent study has demonstrated that self-correction is a powerful tool for code generation, but whether it is effective for smaller models remains unexplored. |
| Approach: | They propose a method that trains small language models to maintain correct outputs while progressively correcting incorrect outputs as turns proceed. |
| Outcome: | The proposed approach improves the ability of small language models for multi-turn code correction. |
ToxReason: A Benchmark for Mechanistic Chemical Toxicity Reasoning via Adverse Outcome Pathway (2026.findings-acl)
Copied to clipboard
| Challenge: | Recent advances in large language models have enabled molecular reasoning for property prediction. however, toxicity arises from complex biological mechanisms, necessitating mechanistic reasoning for reliable prediction. |
| Approach: | They propose a benchmark that evaluates organ-level toxicity reasoning across multiple organs . they find strong predictive performance does not necessarily imply reliable reasoning . |
| Outcome: | The proposed benchmark evaluates toxicity prediction performance and reasoning quality across LLMs. |
CheckEval: A reliable LLM-as-a-Judge framework for evaluating text generation using checklists (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing evaluation protocols for text generation suffer from rating inconsistencies . lexical overlap-based metrics align poorly with human judgments . |
| Approach: | They propose a checklist-based evaluation framework that improves rating reliability via decomposed binary questions. |
| Outcome: | The proposed framework improves rating reliability by decomposing binary questions . it improves agreement across evaluator models by 0.45 and reduces score variance . human evaluation remains the gold standard, but it #, Equal contribution. |
Why So Gullible? Enhancing the Robustness of Retrieval-Augmented Models against Counterfactual Noise (2024.findings-naacl)
Copied to clipboard
| Challenge: | Existing retrieval-augmented language models assume query relevance and irrelevance as dichotomy . existing models are highly brittle to the presence of conflicting information in both the fine-tuning and in-context few-shot learning scenarios. |
| Approach: | They propose methods for handling knowledge conflicts by fine-tuning a discriminator or prompting it to elicit its discriminative capability. |
| Outcome: | The proposed approaches significantly enhance model robustness on open-domain QA. |
Break it Down into BTS: Basic, Tiniest Subword Units for Korean (2022.emnlp-main)
Copied to clipboard
| Challenge: | Existing word embeddings for Korean use the internal structure of words with subword information to improve the quality of word representations. |
| Approach: | They introduce Basic, Tiniest Subword (BTS) units for Korean language that are inspired by Hangeul, the Korean writing system. |
| Outcome: | The proposed framework outperforms the state-of-the-art Korean word embedding by 11.8% on all intrinsic and extrinsic tasks. |
What Changes Can Large-scale Language Models Bring? Intensive Study on HyperCLOVA: Billions-scale Korean Generative Pretrained Transformers (2021.emnlp-main)
Copied to clipboard
Boseop Kim, HyoungSeok Kim, Sang-Woo Lee, Gichang Lee, Donghyun Kwak, Jeon Dong Hyeon, Sunghyun Park, Sungju Kim, Seonhoon Kim, Dongpil Seo, Heungsub Lee, Minyoung Jeong, Sungjae Lee, Minsub Kim, Suk Hyun Ko, Seokhun Kim, Taeyong Park, Jinuk Kim, Soyoung Kang, Na-Hyeon Ryu, Kang Min Yoo, Minsuk Chang, Soobin Suh, Sookyo In, Jinseong Park, Kyungduk Kim, Hiun Kim, Jisu Jeong, Yong Goo Yeo, Donghoon Ham, Dongju Park, Min Young Lee, Jaewook Kang, Inho Kang, Jung-Woo Ha, Woomyoung Park, Nako Sung
| Challenge: | GPT-3 has been used to train large-scale language models on hundreds of billion scale data. |
| Approach: | They propose a Korean variant of GPT-3 that uses Korean tokens to train in-context models. |
| Outcome: | The proposed method shows state-of-the-art zero-shot and few-shot learning on downstream tasks in Korean. |
Leveraging External Knowledge for Historical Document Restoration via Retrieval-Augmented Large Language Models (2026.findings-acl)
Copied to clipboard
| Challenge: | Historical documents suffer from illegibility due to physical deterioration and damage due to deteriorating materials. |
| Approach: | a new framework leverages large language models with retrieval-augmented generation to restore historical documents. authors propose a framework that leverages implicit knowledge of pre-trained LLMs with explicitly retrieved external context. |
| Outcome: | a new framework outperforms existing methods for restoration of historical documents in Korean . the proposed model can restore both general characters and named entities, the authors say . |
Cross-lingual Back-Parsing: Utterance Synthesis from Meaning Representation for Zero-Resource Semantic Parsing (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing approaches to extend semantic parsing (SP) beyond English are challenging due to the complex slot alignment step after translation. |
| Approach: | They propose a method to enhance cross-lingual transfer for SP by utilizing mPLMs. |
| Outcome: | The proposed method synthesizes target language utterances from source meaning representations while maintaining high slot value alignment rates. |
Tree of Clarifications: Answering Ambiguous Questions with Retrieval-Augmented Large Language Models (2023.emnlp-main)
Copied to clipboard
| Challenge: | In open-domain question answering, users often ask ambiguous questions (AQs) . one approach is to identify all possible interpretations of the AQ and generate a long-form answer addressing them all. |
| Approach: | They propose a framework that generates a long-form answer addressing all possible interpretations of an ambiguous question. |
| Outcome: | The proposed framework outperforms baselines on ASQA in a few-shot setup across metrics while surpassing fully-supervised baselines trained on the whole training set in terms of Disambig-F1 and Disambigo-ROUGE. |
You Truly Understand What I Need : Intellectual and Friendly Dialog Agents grounding Persona and Knowledge (2022.findings-emnlp)
Copied to clipboard
Jungwoo Lim, Myunghoon Kang, Yuna Hur, Seung Won Jeong, Jinsung Kim, Yoonna Jang, Dongyub Lee, Hyesung Ji, DongHoon Shin, Seungryong Kim, Heuiseok Lim
| Challenge: | Existing models that ground knowledge and persona at the same time are limited, leading to hallucination and a passive way of using personas. |
| Approach: | They propose a conversational agent that grounds external knowledge and persona simultaneously and a retrieval augmented generation model that generates utterances with lesser hallucination and more engagingness. |
| Outcome: | The proposed agent generates the utterance with lesser hallucination and more engagingness utilizing retrieval augmented generation with knowledge-persona enhanced query. |
SISER: Semantic-Infused Selective Graph Reasoning for Fact Verification (2022.coling-1)
Copied to clipboard
| Challenge: | Existing graph-based methods for fact verification use semantic graphs, which are based on evidence sentences. |
| Approach: | They propose to use semantic-level graph reasoning to inject its reasoning-enhanced representation into other graph-based and sequence-based reasoning methods. |
| Outcome: | The proposed method outperforms the previous graph-based methods and achieves state-of-the-art performance on a large-scale dataset for Fact Extraction and VERification (FEVER). |
A Practical Approach for Building Production-Grade Conversational Agents with Workflow Graphs (2025.acl-industry)
Copied to clipboard
Chiwan Park, Wonjun Jang, Daeryong Kim, Aelim Ahn, Kichang Yang, Woosung Hwang, Jihyeon Roh, Hyerin Park, Hyosun Wang, Min Seok Kim, Jihoon Kang
| Challenge: | Large Language Models (LLMs) have led to significant improvements in various service domains, including search, recommendation, and chatbot applications. |
| Approach: | They propose a framework for developing scalable, controllable, and reliable AI-driven agents that can be applied to real-world applications. |
| Outcome: | The proposed framework bridges the gap between academic research and real-world application, and enables scalable, controllable, and reliable AI-driven agents. |
Exploration-Driven Reinforcement Learning for Expert Routing Improvement in Mixture-of-Experts Language Models (2025.findings-emnlp)
Copied to clipboard
| Challenge: | MoE-based LLMs are not explicitly supervised to select suitable experts. |
| Approach: | They propose Exploration-Driven Reinforcement Learning (ERL) which explicitly optimizes the router by exploration of alternative routing paths. |
| Outcome: | The proposed method improves summarization (SAMSum, XSUM, question answering, and language modeling), and raises routing quality, delivering 8.9 higher MRR than baselines over 100 perturbed routing paths. |
Consistency Training with Virtual Adversarial Discrete Perturbation (2022.naacl-main)
Copied to clipboard
| Challenge: | Existing methods for regularizing a model are agnostic to the training model and may not be effective for perturbed inputs. |
| Approach: | They propose an augmentation method of adding a discrete noise that would incur the highest divergence between predictions by replacing tokens while keeping original semantics. |
| Outcome: | The proposed method outperforms baselines on semi-supervised text classification tasks and a robustness benchmark. |
Enhancing Out-of-Distribution Detection in Natural Language Understanding via Implicit Layer Ensemble (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Out-of-distribution (OOD) detection aims to discern outliers from the intended data distribution, which is crucial to maintaining high reliability and a good user experience. |
| Approach: | They propose a framework that encourages intermediate features to learn layer-specialized representations and assembles them implicitly into a single representation to absorb rich information in the pre-trained language model. |
| Outcome: | The proposed framework is significantly more effective than previous studies in intent classification and OOD datasets. |
Learning from Negative Samples in Biomedical Generative Entity Linking (2025.findings-acl)
Copied to clipboard
| Challenge: | Generative models are usually trained only with positive samples and do not explicitly learn from hard negative samples, which are entities that look similar but have different meanings. |
| Approach: | They propose a framework that trains generative BioEL models using negative samples to learn from hard negative samples. |
| Outcome: | The proposed framework outperforms baseline models by up to an average top-1 accuracy of 1.4% on five benchmarks. |
TA-GRPO-d: Trajectory-Aware GRPO for Optimizing Denoising Trajectories in Diffusion LLMs (2026.acl-long)
Copied to clipboard
| Challenge: | Existing dLLMs rely on fixed denoising schedules and cannot learn efficient unmasking orders. |
| Approach: | They propose a framework that transforms dLLM decoding into a trajectory-aware policy . it uses a confidence-gated denoising strategy that decides which tokens to unmask . |
| Outcome: | The proposed model can learn which tokens to unmask and how many to unmak per step . it can learn the output quality and efficiency of the decoding path itself . |
Enhanced Simultaneous Machine Translation with Word-level Policies (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing studies assume that operations are carried out at the subword level . a novel policy dictates whether to READ or WRITE at each step of the translation process . |
| Approach: | They propose a method to boost SiMT models using language models to address subword disparity . they propose implementing a word-level policy that dictates whether to READ or WRITE . |
| Outcome: | The proposed policy improves the performance of SiMT models by boosting them with language models . the proposed policy plays a vital role in addressing the subword disparity between LMs and SiMT systems. |
Cluster-Guided Label Generation in Extreme Multi-Label Classification (2023.eacl-main)
Copied to clipboard
| Challenge: | Existing classification-based models are poorly per-form for tail labels and ignore semantic relations among labels. |
| Approach: | They propose to guide label generation using label cluster information to hierarchically generate lower-level labels. |
| Outcome: | The proposed model outperforms classification and generation baselines on tail labels and improves in four popular XMC benchmarks. |
Improving Iterative Text Revision by Learning Where to Edit from Other Revision Tasks (2022.emnlp-main)
Copied to clipboard
| Challenge: | Iterative text revision improves text quality by fixing grammatical errors, rephrasing for better readability or contextual appropriateness. |
| Approach: | They propose to build an end-to-end text revision system that can iteratively generate helpful edits by explicitly detecting editable spans with their corresponding edit intents. |
| Outcome: | The proposed system outperforms baselines on other text revision tasks and human evaluations. |
Sound of Story: Multi-modal Storytelling with Audio (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing studies on storytelling with sound have focused on visuals and sounds, but little attention has been given to sound. |
| Approach: | They propose to establish a new component called background sound which is story context-based audio without any linguistic information. |
| Outcome: | The proposed dataset is the largest well-curated dataset for storytelling with sound . it contains 27,354 stories with 19.6 images per story and 984 hours of speech-decoupled audio . |
SafeSearch: Do Not Trade Safety for Utility in LLM Search Agents (2026.findings-eacl)
Copied to clipboard
| Challenge: | Large language model (LLM) based search agents are more likely to produce harmful outputs than base models. |
| Approach: | They propose a query-level shaping term that rewards safe queries and penalizes unsafe ones. |
| Outcome: | The proposed approach reduces harmfulness by over 70% across three red-teaming datasets while producing safe, helpful responses. |
PsyProbe: Proactive and Interpretable Dialogue through User State Modeling for Exploratory Counseling (2026.findings-eacl)
Copied to clipboard
| Challenge: | Existing approaches to mental health dialogue are reactive and lack systematic user state modeling for proactive therapeutic exploration. |
| Approach: | They propose a dialogue system designed for the exploration phase of counseling that systematically tracks user psychological states through the PPPPPI framework augmented with cognitive error detection. |
| Outcome: | The proposed system outperforms baseline and ablation modes in automatic evaluation and expert evaluation by a certified counselor. |
When Thoughts Meet Facts: Reusable Reasoning for Long-Context LMs (2026.findings-acl)
Copied to clipboard
| Challenge: | Recent Long-Context Language Models (LCLMs) do not capture how evidence should be connected . a new framework that integrates thought templates into LCLM frameworks is proving useful . |
| Approach: | They propose a framework that iteratively refines reusable reasoning patterns derived from prior problem solving to improve their templates. |
| Outcome: | The proposed framework outperforms baselines on knowledge-intensive multi-hop reasoning benchmarks and practical scenarios without retrieval. |
When Format Changes Meaning: Investigating Semantic Inconsistency of Large Language Models (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models are vulnerable to semantic inconsistency, a study finds . minor formatting variations result in divergent predictions for semantically equivalent inputs. |
| Approach: | They evaluate LLMs for semantic inconsistency and find they remain vulnerable . they propose to use mechanistic analysis to develop models that improve their reliability . |
| Outcome: | The proposed model is vulnerable to semantic inconsistency, the authors show . their model is brittle even in state-of-the-art models, they say . |
Benchmarking Cognitive Biases in Large Language Models as Evaluators (2024.findings-acl)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have been shown to be effective as automatic evaluators with simple prompting and in-context learning. |
| Approach: | They assemble 16 Large Language Models and evaluate their outputs by preference ranking . they introduce a cognitive bias benchmark to measure six different cognitive biases in LLM evaluation outputs. |
| Outcome: | The proposed model is biased on the CoBBLer benchmark, indicating that machine preferences are misaligned with humans. |
Cactus: Towards Psychological Counseling Conversations using Cognitive Behavioral Theory (2024.findings-emnlp)
Copied to clipboard
Suyeon Lee, Sunghwan Kim, Minju Kim, Dongjin Kang, Dongil Yang, Harim Kim, Minseok Kang, Dayi Jung, Min Kim, Seungbeen Lee, Kyong-Mee Chung, Youngjae Yu, Dongha Lee, Jinyoung Yeo
| Challenge: | Existing models that use large language models are not available due to ethical concerns, and data privacy concerns are a concern. |
| Approach: | They propose a multi-turn dialogue dataset that emulates real-life counseling interactions using the goal-oriented approach of Cognitive Behavioral Therapy (CBT). |
| Outcome: | The proposed model outperforms other models in counseling skills, highlighting its effectiveness and potential as a counseling agent. |
DaCoM: Strategies to Construct Domain-specific Low-resource Language Machine Translation Dataset (2025.coling-industry)
Copied to clipboard
| Challenge: | Existing models for low-resource languages struggle with domain-specific terms and lack of expert annotators for dataset creation. |
| Approach: | They propose a method for collecting low-resource language pairs from industrial domains using a large language model and neural machine translation framework. |
| Outcome: | The proposed model performs poorly on DaCoM-created datasets with up to 53.7 BLEURT points difference depending on domain inclusion. |
3D-Aware Vision-Language Models Fine-Tuning with Geometric Distillation (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Vision-Language Models (VLMs) have shown remarkable performance on diverse visual and linguistic tasks, yet they remain limited in their understanding of 3D spatial structures. |
| Approach: | They propose a framework that injects human-inspired geometric cues into pretrained VLMs . they use sparse correspondences, relative depth relations and dense cost volumes . |
| Outcome: | The proposed framework outperforms existing methods on vision-language reasoning and 3D perception benchmarks. |
CookingSense: A Culinary Knowledgebase with Multidisciplinary Assertions (2024.lrec-main)
Copied to clipboard
| Challenge: | CookingSense is a descriptive collection of knowledge assertions in the culinary domain extracted from various sources, including web data, scientific papers, and recipes. |
| Approach: | They introduce CookingSense, a descriptive collection of knowledge assertions in the culinary domain extracted from various sources, including web data, scientific papers, and recipes. |
| Outcome: | The proposed system improves retrieval augmented language models and food decision support systems. |
Simple Questions Generate Named Entity Recognition Datasets (2022.emnlp-main)
Copied to clipboard
| Challenge: | Recent named entity recognition models rely on human-annotated datasets . however, in-domain dictionaries and sentences are often unavailable or expensive to construct for many entity types. |
| Approach: | They propose an ask-to-generate approach which automatically generates NER datasets by asking natural language questions to an open-domain question answering system. |
| Outcome: | The proposed model outperforms the previous best model by 19.5 F1 score on six benchmarks and achieves state-of-the-art performance. |
MiLQ: Benchmarking IR Models for Bilingual Web Search with Mixed Language Queries (2025.emnlp-main)
Copied to clipboard
| Challenge: | Information Retrieval (IR) research on mixed-language queries remains sparse and outdated. |
| Approach: | They propose a test set for mixed-language queries that is realistic and preferred by bilingual speakers. |
| Outcome: | The proposed benchmarks show that multilingual IR models perform moderately on MiLQ and inconsistently across native, English, and mixed-language queries. |
From Relevance to Authority: Authority-aware Generative Retrieval in Web Search Engines (2026.acl-industry)
Copied to clipboard
| Challenge: | Existing methods that optimize for relevance overlook document trustworthiness . Generative information retrieval (GenIR) is a promising paradigm for retrieval tasks . |
| Approach: | They propose an Authority-aware Generative Retriever (AuthGR) that incorporates authority into GenIR. |
| Outcome: | The proposed framework improves authority and accuracy in real-world user engagement and reliability. |
Assessing LLM Reasoning Steps via Principal Knowledge Grounding (2025.findings-emnlp)
Copied to clipboard
Hyeon Hwang, Yewon Cho, Chanwoong Yoon, Yein Park, Minju Song, Kyungjae Lee, Gangwoo Kim, Jaewoo Kang
| Challenge: | Step-by-step reasoning has become a standard approach for large language models to tackle complex tasks. |
| Approach: | They propose a framework that assesses the knowledge grounding of intermediate reasoning by using a large-scale repository of atomic knowledge essential for reasoning. |
| Outcome: | The evaluation suite identifies missing or misapplied knowledge elements and provides crucial insights for uncovering fundamental reasoning deficiencies in LLMs. |
Mismatch between Multi-turn Dialogue and its Evaluation Metric in Dialogue State Tracking (2022.acl-short)
Copied to clipboard
| Challenge: | Existing evaluation metrics for dialog state tracking are limited for belief states accumulated as dialog proceeds . relative slot accuracy allows intuitive evaluation by assigning relative scores according to the turn of each dialog . |
| Approach: | They propose to use relative slot accuracy to complement existing evaluation metrics . joint goal accuracy and slot accuracy are used to evaluate accumulated belief states . |
| Outcome: | The proposed metrics focus on "penalizing states that fail to predict," not "reward for well-predicted states" the proposed metrics do not depend on the number of predefined slots, and allow intuitive evaluation . |
Leveraging Order-Free Tag Relations for Context-Aware Recommendation (2021.emnlp-main)
Copied to clipboard
| Challenge: | Existing approaches to tag recommendation neglect orderlessness and inter-dependency . Empirical results on Instagram and Stack Overflow show that our method is significantly superior to the previous approaches. |
| Approach: | They propose a sequence-oblivious generation method for tag recommendation . the next tag to be generated is independent of the order of the generated tags . they also propose regressive generation methods that take orderlessness into account . |
| Outcome: | Empirical results show that the proposed method is superior to previous approaches . the proposed system is based on two domains, Instagram and Stack Overflow . |
Understanding Iterative Revision from Human-Written Text (2022.acl-long)
Copied to clipboard
| Challenge: | This work describes IteraTeR: the first large-scale, multi-domain, edit-intention annotated corpus of iteratively revised text. |
| Approach: | They propose to annotate iteratively revised text using a multi-domain annotated corpus that generalizes to a variety of domains, edit intentions, revision depths, and granularities. |
| Outcome: | The proposed model improves automatic evaluations by integrating edit intentions with writing quality. |
STAGE: Simple Text Data Augmentation by Graph Exploration (2024.lrec-main)
Copied to clipboard
| Challenge: | Pre-trained language models (PLMs) are widely used for various tasks, but fine-tuning them requires sufficient data. |
| Approach: | They propose a method for data augmentation that utilizes a word-relation graph to select optimal words for each modification. |
| Outcome: | The proposed method is highly effective across diverse datasets and different PLMs. |
Graph-Induced Transformers for Efficient Multi-Hop Question Answering (2022.emnlp-main)
Copied to clipboard
| Challenge: | Recent MHQA tasks that require inter-paragraph/sentence linkages use graphs to model internal structural information within text. |
| Approach: | They propose a graph-induced transformer that applies graph-derived attention patterns directly into a PLM without external graph modules. |
| Outcome: | The proposed model can replace external graph modules while preserving model performance. |
Unifying Uniform and Binary-coding Quantization for Accurate Compression of Large Language Models (2025.acl-long)
Copied to clipboard
Seungcheol Park, Jeongin Bae, Beomseok Kwon, Minjun Kim, Byeongwook Kim, Se Jung Kwon, U Kang, Dongsoo Lee
| Challenge: | Quantization is essential for deploying large language models (LLMs) efficiently since they require expensive computational and memory costs. |
| Approach: | They propose a quantization method that unifies flexible mapping techniques to optimize parameters precisely. |
| Outcome: | The proposed method outperforms existing methods and achieves higher accuracy on GSM8K benchmark. |
Becoming Experienced Judges: Selective Test-Time Learning for Evaluators (2026.eacl-short)
Copied to clipboard
| Challenge: | Large language models and visionlanguage models are increasingly used as automatic evaluators. |
| Approach: | They propose a framework that allows evaluators to improve *sequentially* at inference time without additional training or external signals. |
| Outcome: | The proposed framework outperforms strong baselines in two pairwise comparisons. |
Subject-level Inference for Realistic Text Anonymization Evaluation (2026.acl-long)
Copied to clipboard
Myeong Seok Oh, Dong-Yun Kim, Hanseok Oh, Chaean Kang, Joeun Kang, Xiaonan Wang, Hyunjung Park, Young Cheol Jung, Hansaem Kim
| Challenge: | Existing text anonymization evaluations assume only a single data subject, ignoring multi-subject scenarios. |
| Approach: | They propose a benchmark that shifts the unit of evaluation from text spans to individuals . they show that subject-level inference protection drops as low as 33% when masked . |
| Outcome: | The proposed benchmark reduces the amount of protection available when PII spans are masked. |
K-MetBench: A Multi-Dimensional Benchmark for Fine-Grained Evaluation of Expert Reasoning, Locality, and Multimodality in Meteorology (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing evaluation frameworks for large language models for domain specific tasks are coarse and do not provide a multidimensional evaluation of a model's ability to interpret domain specific data. |
| Approach: | They propose a diagnostic benchmark grounded in national qualification exams that exposes critical gaps across four dimensions: expert visual reasoning of charts, logical validity via expert-verified rationales, Korean-specific geo-cultural comprehension, and fine-grained domain analysis. |
| Outcome: | The proposed model outperforms global models in local contexts, demonstrating that parameter scaling alone cannot resolve cultural dependencies. |
Jailbreaking Multimodal Large Language Models using Multi-Clip Video (2026.acl-long)
Copied to clipboard
| Challenge: | Existing studies show that video inputs can bypass safety alignment, yet it remains unclear which properties of video input induce this vulnerability. |
| Approach: | They propose a simple image-based defense that mitigates the vulnerability of MLLMs by analyzing video inputs. |
| Outcome: | The proposed defense leverages the relative robustness of the image modality. |
Dialogue Chain-of-Thought Distillation for Commonsense-aware Conversational Agents (2023.emnlp-main)
Copied to clipboard
Hyungjoo Chae, Yongho Song, Kai Ong, Taeyoon Kwon, Minjin Kim, Youngjae Yu, Dongha Lee, Dongyeop Kang, Jinyoung Yeo
| Challenge: | a human-like chatbot requires commonsense reasoning to comprehend and respond to information . however, identifying and aggregating key evidence within a single hop is a challenge . a knowledge distillation framework is proposed that leverages LLMs as unreliable teachers . |
| Approach: | They propose a framework that leverages large language models as unreliable teachers to facilitate multi-hop reasoning over a dialogue context. |
| Outcome: | The proposed framework leverages LLMs as unreliable teachers and selectively distills consistent and helpful rationales via alignment filters. |
Keep Me Updated! Memory Management in Long-term Conversations (2022.findings-emnlp)
Copied to clipboard
Sanghwan Bae, Donghyun Kwak, Soyoung Kang, Min Young Lee, Sungdong Kim, Yuin Jeong, Hyeri Kim, Sang-Woo Lee, Woomyoung Park, Nako Sung
| Challenge: | Existing studies do not deal with cases where memorized information is outdated, which may cause confusion in later conversations. |
| Approach: | They propose a task where bots keep track of and bring up the latest information about users while conversing through multiple sessions. |
| Outcome: | The proposed method outperforms baselines that leave the stored memory unchanged in terms of engagingness and humanness, and a larger performance gap in the later sessions. |
Fine-tuning CLIP Text Encoders with Two-step Paraphrasing (2024.findings-eacl)
Copied to clipboard
| Challenge: | Contrastive language-image pre-training models have demonstrated considerable success across various vision-language tasks, such as text-to-image retrieval. |
| Approach: | They propose a fine-tuning approach to enhance the representations of CLIP models for paraphrases by leveraging large language models. |
| Outcome: | The proposed model improves on baseline models across paraphrased retrieval, visual genome relation and attribution, and seven semantic textual similarity tasks. |
Tracing Mathematical Proficiency Through Problem-Solving Processes (2026.findings-acl)
Copied to clipboard
| Challenge: | Knowledge Tracing (KT) models a learner's evolving knowledge state over time, but lacks the rich information embedded in students' problem-solving processes. |
| Approach: | They propose a framework that uses a teacher-student-teacher pipeline to extract students’ Mathematical Proficiency (MP) as intermediate representation. |
| Outcome: | The proposed framework improves the prediction performance of existing KT methods and provides interpretable explanations by explicitly modeling students’ mathematical proficiency. |
Beyond Candidates : Adaptive Dialogue Agent Utilizing Persona and Knowledge (2023.findings-emnlp)
Copied to clipboard
| Challenge: | a previous study suggested that human dialogue systems ground persona and knowledge but they require incomplete candidate sets. |
| Approach: | They propose an adaptive dialogue agent that uses persona and knowledge without candidate sets . their model generates consistent and relevant persona descriptions and identifies relevant knowledge . |
| Outcome: | The proposed model outperforms baselines that ground persona and knowledge candidates even with fragmentary information. |
C2: Scalable Auto-Feedback for LLM-based Chart Generation (2025.naacl-long)
Copied to clipboard
Woosung Koh, Janghan Yoon, MinHyung Lee, Youngjin Song, Jaegwan Cho, Jaehyun Kang, Taehyeon Kim, Se-Young Yun, Youngjae Yu, Bongshin Lee
| Challenge: | generating high-quality charts with Large Language Models presents significant challenges due to limited data and the high cost of curation. |
| Approach: | They propose a referencefree automatic feedback generator to generate high-quality charts with Large Language Models. |
| Outcome: | The proposed framework outperforms baselines and shows that it significantly improves data diversity. |
“Killing Me” Is Not a Spoiler: Spoiler Detection Model using Graph Neural Networks with Dependency Relation-Aware Attention Mechanism (2021.eacl-main)
Copied to clipboard
| Challenge: | Several attention-based spoiler detection models are insufficient for utilizing dependency relations between context words. |
| Approach: | They propose a new spoiler detection model called SDGNN that uses syntax-aware graph neural networks to detect dependency relations between context words. |
| Outcome: | The proposed model outperforms existing models on two real-world benchmark datasets. |
X-SNS: Cross-Lingual Transfer Prediction through Sub-Network Similarity (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Cross-lingual transfer (XLT) is an emergent ability of multilingual language models that preserves their performance when evaluated in non-English languages. |
| Approach: | They propose to use sub-network similarity between two languages as a proxy for XLT prediction. |
| Outcome: | The proposed method shows proficiency in ranking candidates for zero-shot XLT, achieving an improvement of 4.6% on average in terms of NDCG@3. |
Can You Trick the Grader? Adversarial Persuasion of LLM Judges (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models (LLMs) are increasingly used as automated evaluators in practical settings . |
| Approach: | a study by the university of california reveals that persuasive language can bias large language models when scoring mathematical reasoning tasks. |
| Outcome: | The proposed model can bias judges when scoring mathematical reasoning tasks . Consistency causes the most severe distortion, with Consistencies leading to 8% distortion . |
SCRIPTMIND: Crime Script Inference and Cognitive Evaluation for LLM-based Social Engineering Scam Detection System (2026.eacl-industry)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have shown promise in identifying deception, but their cognitive assistance potential remains underexplored. |
| Approach: | They propose a framework for LLM-based scam detection that bridges automated reasoning and human cognition. |
| Outcome: | The proposed framework outperforms GPT-4o in the Korean scam detection and phone scam simulations. |
PRINCIPLES: Synthetic Strategy Memory for Proactive Dialogue Agents (2025.findings-emnlp)
Copied to clipboard
Namyoung Kim, Kai Tzu-iunn Ong, Yeonjun Hwang, Minseok Kang, Iiseo Jihn, Gayoung Kim, Minju Kim, Jinyoung Yeo
| Challenge: | Existing strategies for proactive dialogue face limitations such as limited strategy coverage and preference bias in planning. |
| Approach: | They propose a synthetic strategy memory for proactive dialogue agents based on large language models . PRINCIPLES is derived through offline self-play simulations and serves as reusable knowledge that guides strategy planning during inference. |
| Outcome: | PRINCIPLES is a synthetic strategy memory for proactive dialogue agents. |
Boosting Prompt-Based Self-Training With Mapping-Free Automatic Verbalizer for Multi-Class Classification (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Recent prompt-based fine-tuning techniques have garnered considerable interest as a core technique for few-shot text classification tasks. |
| Approach: | They propose a prompt-based fine-tuning approach that reformulates the fine-uning objective to align with the Masked Language Modeling objective. |
| Outcome: | The proposed method has shown superior performance on five multi-class classification datasets. |
Denoising Table-Text Retrieval for Open-Domain Question Answering (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing studies in table-text open-domain question answering have problems with false-positive labels in training datasets. |
| Approach: | They propose a denoised table-text retriever that discards false positives from training datasets . they integrate table-level ranking information into the retriever to assist in finding evidence . |
| Outcome: | The proposed method outperforms baselines on retrieval recall and QA tasks. |
Self-Consistent Reasoning-based Aspect-Sentiment Quad Prediction with Extract-Then-Assign Strategy (2024.findings-acl)
Copied to clipboard
| Challenge: | Recent studies have developed powerful generative methods for aspect sentiment quad prediction (ASQP) but they still suffer from imprecise predictions and limited interpretability due to data scarcity and inadequate modeling of the quadruplet composition process. |
| Approach: | They propose a self-consistent reasoning-based aspect sentiment quadruple prediction framework which generates reasonings and corresponding quadruples in sequence. |
| Outcome: | The proposed model significantly improves its ability to handle complex reasoning tasks and correctly predict quadruplets through consistency voting, resulting in enhanced interpretability and accuracy in aspect sentiment quadr uplp prediction. |
Can You Share Your Story? Modeling Clients’ Metacognition and Openness for LLM Therapist Evaluation (2025.findings-acl)
Copied to clipboard
Minju Kim, Dongje Yoo, Yeonjun Hwang, Minseok Kang, Namyoung Kim, Minju Gwak, Beong-woo Kwak, Hyungjoo Chae, Harim Kim, Yunjoong Lee, Min Hee Kim, Dayi Jung, Kyong-Mee Chung, Jinyoung Yeo
| Challenge: | Existing evaluation methods for psychological counseling rely on client simulators that clearly disclose internal states to the therapist, making it difficult to determine whether an LLM therapist can uncover unexpressed perspectives. |
| Approach: | They propose a new evaluation framework featuring a controllable and realistic client simulator which dynamically adapts itself based on the ongoing counseling session. |
| Outcome: | The proposed evaluation framework features a realistic and controllable client simulator which dynamically adapts itself based on the ongoing counseling session, offering a more realistic and challenging evaluation environment. |
FractalLLM: Lossless Self-Speculative Decoding with Layer Embedded Self-Compression (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Autoregressive decoding requires a full forward pass for each generated token, increasing inference latency. |
| Approach: | They propose a lossless self-speculative decoding method that embeds a compressed model within selected decoder layers of the original model. |
| Outcome: | The proposed method achieves substantial speed-ups (up to 2.47) over standard autoregressive decoding. |
Threads of Subtlety: Detecting Machine-Generated Texts Through Discourse Motifs (2024.acl-long)
Copied to clipboard
| Challenge: | Empirical findings show that although both LLMs and humans generate distinct discourse patterns influenced by specific domains, human-written texts exhibit more structural variability, reflecting the nuanced nature of human writing in different domains. |
| Approach: | They propose a method to leverage hierarchical parse trees and recursive hypergraphs to uncover distinctive discourse patterns in texts written by humans and LLMs. |
| Outcome: | The proposed method combines hierarchical parse trees and recursive hypergraphs to uncover distinctive discourse patterns in texts produced by both LLMs and humans. |
ToolHaystack: Stress-Testing Tool-Augmented Language Models in Realistic Long-Term Interactions (2025.findings-emnlp)
Copied to clipboard
Beong-woo Kwak, Minju Kim, Dongha Lim, Hyungjoo Chae, Dongjin Kang, Sunghwan Kim, Dongil Yang, Jinyoung Yeo
| Challenge: | Existing evaluations assume tool use in short contexts, offering limited insight into model behavior during realistic long-term interactions. |
| Approach: | a benchmark is a tool to test long-term tool use in large language models . the tool includes multiple tasks execution contexts and realistic noise . |
| Outcome: | a new benchmark tests the tool use capabilities in long-term interactions. |
Med-PRM: Medical Reasoning Models with Stepwise, Guideline-verified Process Rewards (2025.emnlp-main)
Copied to clipboard
Jaehoon Yun, Jiwoong Sohn, Jungwoo Park, Hyunjae Kim, Xiangru Tang, Daniel Shao, Yong Hoe Koo, Ko Minhyeok, Qingyu Chen, Mark Gerstein, Michael Moor, Jaewoo Kang
| Challenge: | Large language models have shown promise in clinical decision making, but current approaches struggle to localize and correct reasoning errors at specific steps of the reasoning process. |
| Approach: | They propose a process reward modeling framework that leverages retrieval-augmented generation to verify each reasoning step against established medical knowledge bases. |
| Outcome: | The proposed model improves on five medical QA benchmarks and two open-ended diagnostic tasks by 13.50% on MedQA. |
GenPTQ: Green Post-Training Quantization for Large-Scale ASR Models with Mixed-Precision Bit Allocation (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Large-scale models with high memory and computation demands face significant challenges for deployment. |
| Approach: | They propose a weight-only quantization method that optimizes for weights and QAT to restore accuracy. |
| Outcome: | The proposed method achieves 89.1% model size reduction with only 0.8% increase in WER and completes optimization in just 15 seconds. |
Reasoning Visual Dialog with Sparse Graph Learning and Knowledge Transfer (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Visual dialog is a task of answering questions grounded in an image using dialog history as context. |
| Approach: | They propose a Sparse Graph Learning method to formulate visual dialog as a graph structure learning task. |
| Outcome: | The proposed model outperforms the state-of-the-art models on the VisDial v1.0 dataset. |
Can Language Models be Biomedical Knowledge Bases? (2021.emnlp-main)
Copied to clipboard
| Challenge: | Existing studies have focused on probing LMs in the general domain but little attention has been given to whether they can be used as domain knowledge bases. |
| Approach: | They propose to use 49K biomedical factual knowledge triples to probe LMs for biomedically . they find that biomedic LM can achieve up to 18.51% Acc@5 on retrieving biomedcial knowledge. |
| Outcome: | The proposed biomedical factual knowledge probing benchmark achieves 18.51% Acc@5 on biomedically-relevant knowledge retrieval. |
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. |
Ask Optimal Questions: Aligning Large Language Models with Retriever’s Preference in Conversation (2025.findings-naacl)
Copied to clipboard
| Challenge: | Existing methods to perform conversational search are sub-optimal due to the limited ability to incorporate signals from the retrieval results. |
| Approach: | They propose to optimize a language model for reformulating search queries in line with retrievers’ preferences by combining a large-scale dataset with Retrievers’ Feedback. |
| Outcome: | The proposed framework outperforms existing methods on two benchmarks and surpasses the state-of-the-art methods. |
LLMs can be easily Confused by Instructional Distractions (2025.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) show exceptional skill in instruction following tasks, but can become vulnerable when they are required to disregard instructions. |
| Approach: | They propose a benchmark to assess LLMs' performance under instructional distraction. |
| Outcome: | The proposed benchmark categorizes real-world instances of instructional distraction and evaluates LLMs across four instruction tasks: proofreading, rewriting, translation, and style transfer—alongside five input tasks: reasoning, code generation, mathematical reasoning, bias detection, and question answering. |
Distilling Cross-Modal Knowledge into Domain-Specific Retrievers for Enhanced Industrial Document Understanding (2025.emnlp-industry)
Copied to clipboard
Jinhyeong Lim, Jeongwan Shin, Seeun Lee, Seongdeok Kim, Joungsu Choi, Jongbae Kim, Chun Hwan Jung, Youjin Kang
| Challenge: | Retrieval-Augmented Generation (RAG) has shown strong performance in open-domain tasks, but its effectiveness in industrial domains is limited by a lack of domain understanding and document structural elements (DSE) such as tables, figures, charts, and formula. |
| Approach: | They propose a knowledge distillation framework that transfers complementary knowledge from Large Language Models and Vision-Language Models into a compact domain-specific retriever. |
| Outcome: | The proposed framework outperforms larger baselines while requiring significantly less computational complexity. |
Universal Domain Adaptation for Robust Handling of Distributional Shifts in NLP (2023.findings-emnlp)
Copied to clipboard
Hyuhng Kim, Hyunsoo Cho, Sang-Woo Lee, Junyeob Kim, Choonghyun Park, Sang-goo Lee, Kang Yoo, Taeuk Kim
| Challenge: | Despite advances in computer vision, its application on language input still needs to be explored despite its feasibility. |
| Approach: | They propose a universal domain adaptation (uniDA) benchmark for natural language that offers thorough viewpoints of the model’s generalizability and robustness. |
| Outcome: | The proposed model can handle spoken language in the real world while also detecting unprocessable inputs from the target domain. |
SharVeT: Similarity-aware Parameter Sharing with Vector-based Tuning for Efficient LLM Compression (2026.acl-long)
Copied to clipboard
| Challenge: | Existing methods for parameter sharing rely on naive grouping and fail to correct sharing-induced discrepancies. |
| Approach: | They propose a parameter sharing framework that performs similarity-based grouping to ensure accurate sharing and allocates parameters adaptively to preserve diversity within each group. |
| Outcome: | The proposed framework outperforms existing methods, achieving 32.1% lower perplexity and 23.3% higher few-shot reasoning accuracy. |
Personalized LLM Decoding via Contrasting Personal Preference (2025.emnlp-main)
Copied to clipboard
| Challenge: | Personalization of large language models (LLMs) is becoming increasingly important as they are increasingly deployed in real-world applications. |
| Approach: | They propose a decoding-time approach that leverages the user's implicit reward signal by performing parameter-efficient fine-tuning on user-specific data. |
| Outcome: | The proposed approach improves personalization by an average of 10.57% in ROUGE-L without external reward models or additional training procedures. |
AG-GRPO: Answer-Guided GRPO for Masked Diffusion Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | Recent work on large language models (LLMs) has emphasized not only final-answer accuracy but also reliability of reasoning on challenging tasks. |
| Approach: | They propose an answer-guided group-relative policy optimization for masked diffusion language models which generates text through iterative mangled token restoration. |
| Outcome: | The proposed approach improves over pretrained dLLMs and prior RL methods across mathematics, puzzle-solving, and code-generation benchmarks. |
Building a Role Specified Open-Domain Dialogue System Leveraging Large-Scale Language Models (2022.naacl-main)
Copied to clipboard
| Challenge: | Recent large-scale language models have produced human-like responses in open-domain dialogue systems. |
| Approach: | They propose a framework for imposing roles on open-domain dialogue systems . they use few-shot learning to build a Korean dialogue dataset from scratch . |
| Outcome: | The proposed framework meets role specifications while maintaining conversational abilities. |
DoMIX: An Efficient Framework for Exploiting Domain Knowledge in Fine-Tuning (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for domain-adaptive pre-training (DAP) face several limitations: high computational cost and GPU memory usage during training; and lack of generalized model for all end tasks. |
| Approach: | They propose a domain-adaptive pre-training (DAP) method that uses a representative parameter-efficient fine-tuning method to provide pre-trained models for specific tasks. |
| Outcome: | The proposed method can be extended beyond the DAP setting to standard LLM fine-tuning scenarios. |
infoVerse: A Universal Framework for Dataset Characterization with Multidimensional Meta-information (2023.acl-long)
Copied to clipboard
| Challenge: | Several methods for characterizing datasets based on model-driven meta-information have been developed, but the relationship and complementary effects of these methods have received less attention. |
| Approach: | They propose a framework that captures multidimensional characteristics of datasets by incorporating various model-driven meta-information. |
| Outcome: | The proposed framework outperforms baselines in three real-world applications and can be used in a variety of real-time problems. |
Aligning Large Language Models through Synthetic Feedback (2023.emnlp-main)
Copied to clipboard
| Challenge: | Currently, alignment learning requires significant human demonstrations and feedback from proprietary LLMs such as ChatGPT. |
| Approach: | They propose a framework that uses synthetic feedback to align large language models to human values without extensive human annotations and proprietary LLMs. |
| Outcome: | The proposed model outperforms open-source models on human-annotated demonstrations in alignment benchmarks. |
PLEX: Adaptive Parameter-Efficient Fine-Tuning for Code LLMs using Lottery-Tickets (2025.naacl-industry)
Copied to clipboard
| Challenge: | PLEX is a lottery-ticket based parameter-efficient fine-tuning method that adapts large language models to well-supported and underrepresented programming languages (PLs) in pretraining. |
| Approach: | They propose a lottery-ticket based parameter-efficient fine-tuning method that adapts large language models to well-supported and underrepresented programming languages (PLs) |
| Outcome: | The proposed method achieves state-of-the-art performance among PEFT methods while maintaining competitive results with reduced computational overhead. |
II-MMR: Identifying and Improving Multi-modal Multi-hop Reasoning in Visual Question Answering (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing studies have focused on assessing the model’s overall accuracy without evaluating it on different reasoning cases. |
| Approach: | They propose a novel idea to identify and improve multi-modal multi-hop reasoning in VQA by using two new language prompts to find a reasoning path to reach its answer. |
| Outcome: | The proposed model improves multi-modal multi-hop reasoning in visual question answering (VQA) it finds that the proposed model is easy to answer, simply demanding “single-hop” reasoning, whereas only a few questions require “multi-hop.” |
Rethinking Reward Model Evaluation Through the Lens of Reward Overoptimization (2025.acl-long)
Copied to clipboard
| Challenge: | Existing benchmarks for reward models show a weak correlation with performance of optimized policies . existing benchmarks do not accurately assess the true capabilities of reward models . |
| Approach: | They explore how reward overoptimization captures how well a reward model aligns with human preferences and the dynamics of the learning signal it provides to the policy. |
| Outcome: | The proposed benchmarks show that reward overoptimization is a weak factor . the high correlation with degree of overoptimalization leads to lower correlation with downstream performance . |
Can Large Language Models Differentiate Harmful from Argumentative Essays? Steps Toward Ethical Essay Scoring (2025.coling-main)
Copied to clipboard
| Challenge: | Existing automated essay scoring systems overlook ethical and moral aspects of content, erroneously assigning high scores to essays that propagate harmful opinions. |
| Approach: | They introduce a Harmful Essay Detection benchmark to test the effectiveness of various Large Language Models (LLMs) they find that current AES systems overlook ethically and morally problematic elements in essays . |
| Outcome: | The proposed benchmark compared LLMs and AES models to identify and score harmful essays. |
Look at the First Sentence: Position Bias in Question Answering (2020.emnlp-main)
Copied to clipboard
| Challenge: | Extractive question answering models are trained to predict start and end positions of answers . recent QA models outperform humans in some datasets due to their simplicity and effectiveness. |
| Approach: | They propose to use prior distribution of answer positions as a bias model to reduce position bias. |
| Outcome: | The proposed model outperforms BERT from 37.48% to 81.64% when trained on a biased SQUAD dataset. |