Papers by Jiho Lee

7 papers
Speculative Verification: Exploiting Information Gain for Speculative Decoding (2026.findings-acl)

Copied to clipboard

Challenge: Large Language Models (LLMs) are used for many applications but their size and computational cost make inference serving a significant challenge.
Approach: They propose an efficient augmentation to Speculative Decoding (SD) that predicts speculation accuracy and dynamically adapts the verification length to maximize throughput.
Outcome: The proposed model reduces wasted verification on rejected tokens and improves decoding efficiency.
SuperST: Superficial Self-Training for Few-Shot Text Classification (2024.lrec-main)

Copied to clipboard

Challenge: In few-shot text classification, self-training relies on pseudo-labels to expand data, which has shown success, but can accumulate errors due to noisy pseudo-labeled data.
Approach: They propose a method to mitigate noise in noisy pseudo-labeled data by applying superficial learning to noisy data and fine-tuning to less noisy data.
Outcome: The proposed framework improves the classifier accuracy for few-shot text classification by 18.5% at most and 8% in average, compared with the state-of-the-art SSL baselines.
FINEST: Improving LLM Responses to Sensitive Topics Through Fine-Grained Evaluation (2026.findings-eacl)

Copied to clipboard

Challenge: Existing evaluation frameworks lack systematic methods to identify weaknesses in LLMs . Existing methods to evaluate LLM responses to sensitive topics are lacking .
Approach: They propose a FINE-grained response evaluation taxonomy for sensitive topics that breaks down helpfulness and harmlessness into errors across three main categories: Content, Logic, and Appropriateness.
Outcome: The proposed model outperforms refinement without guidance on Korean-sensitive questions . FINEST significantly improves the model responses across all three categories .
Exploring Cross-Cultural Differences in English Hate Speech Annotations: From Dataset Construction to Analysis (2024.naacl-long)

Copied to clipboard

Challenge: Existing datasets for hate speech detection neglect the cultural diversity within a single language.
Approach: They propose a CR**oss-cultural **E**nglish **Hate* speech dataset that uses culturally hateful keywords to identify posts from four countries plus the United States.
Outcome: The proposed dataset shows that only 56.2% of the posts in CREHate achieve consensus among all countries, with the highest pairwise label difference rate of 26%.
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.
Better Quality Estimation for Low Resource Corpus Mining (2022.findings-acl)

Copied to clipboard

Challenge: State-of-the-art Quality Estimation models lack robustness to out-of domain examples.
Approach: They propose a method that uses multitask training, data augmentation and contrastive learning to achieve better and more robust QE performance.
Outcome: The proposed method improves QE performance significantly in the MLQE challenge and the robustness of QE models when tested in the Parallel Corpus Mining setup.
Two-Step Question Retrieval for Open-Domain QA (2022.findings-acl)

Copied to clipboard

Challenge: Existing question retrieval models have shown a significant increase in inference speed but at the cost of lower QA performance compared to the retriever-reader pipeline.
Approach: They propose a two-step question retrieval model with distant supervision to improve inference speed.
Outcome: The proposed model significantly increases the performance of existing question retrieval models with a negligible loss on inference speed.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations