Papers by Shinwoo Park
DITTO: A Spoofing Attack Framework on Watermarked LLMs via Knowledge Distillation (2026.eacl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) generate coherent, human-like text at scale, but raises concerns about authenticity and trust. |
| Approach: | They propose a threat of watermark spoofing that allows a malicious model to generate text containing the authentic-looking watermark of a trusted, victim model. |
| Outcome: | The proposed attack repurposes watermark radioactivity from a discoverable trait into an attack vector and replicates it. |
KatFishNet: Detecting LLM-Generated Korean Text through Linguistic Feature Analysis (2025.acl-long)
Copied to clipboard
| Challenge: | Detecting LLM-generated text is crucial for academic integrity, preventing plagiarism, protecting copyrights, ethical research practices. |
| Approach: | They propose a method specifically designed for Korean language to detect LLM-generated text . they examine spacing patterns, part-of-speech diversity, and comma usage . |
| Outcome: | The proposed method achieves an average of 19.78% higher AUC-ROC compared to the best-performing detection method. |
Generalizable Implicit Hate Speech Detection Using Contrastive Learning (2022.coling-1)
Copied to clipboard
| Challenge: | Hate speech detection is challenging when there are insufficient lexical cues. |
| Approach: | They propose a contrastive learning method that pulls an implication and its corresponding posts close in representation space. |
| Outcome: | The proposed method improves on BERT and HateBERT benchmarks on three implicit hate speech benchmarks. |
A Linguistics-Aware LLM Watermarking via Syntactic Predictability (2026.acl-long)
Copied to clipboard
| Challenge: | a central challenge remains balancing text quality against detection robustness. |
| Approach: | They propose a framework that aligns watermark strength with linguistic degrees of freedom . they use part-of-speech models to weaken the signal in grammatically constrained contexts . |
| Outcome: | The proposed framework outperforms existing methods in linguistic indeterminacy tests on languages . it weakens the watermark strength in grammatically constrained contexts and strengthens it in contexts with greater linguistic flexibility. |
TrapDoc: Deceiving LLM Users by Injecting Imperceptible Phantom Tokens into Documents (2025.findings-emnlp)
Copied to clipboard
| Challenge: | a growing number of LLMs have been used to provide reasoning, writing, text-editing capabilities. |
| Approach: | They propose a method to inject imperceptible phantom tokens into LLMs to deceive users . the technique generates outputs that appear plausible to users but are in fact incorrect . |
| Outcome: | a new method injects imperceptible tokens into documents to deceive users . the proposed framework is compared to baselines to show its effectiveness . |
Contrastive Learning with Keyword-based Data Augmentation for Code Search and Code Question Answering (2023.eacl-main)
Copied to clipboard
| Challenge: | Recent work on code search proposes data augmentation of queries for contrastive learning. |
| Approach: | They propose to augment query-code pairs with key words to preserve key words . they use keyDAC to fine-tune various pre-trained language models . |
| Outcome: | The proposed approach outperforms the current state-of-the-art in code search and question answering tasks. |
Mondrian: A Framework for Logical Abstract (Re)Structuring (2025.emnlp-main)
Copied to clipboard
| Challenge: | Using the ABT structure, academic abstracts are structured to provide clear and concise prose, but a lack of clarity and logical coherence is a challenge for authors struggling with English proficiency or academic writing conventions. |
| Approach: | They propose a framework that identifies the key components of an abstract and reorients itself to properly reflect the ABT logical progression. |
| Outcome: | The proposed framework improves comprehensibility of academic writing, particularly for non-native English speakers, and is based on a human evaluation and automated metrics. |
ConPrompt: Pre-training a Language Model with Machine-Generated Data for Implicit Hate Speech Detection (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Existing pre-trained language models for hate speech detection are not specialized in implicit hate speech. |
| Approach: | They propose a pre-trained language model for implicit hate speech detection that leverages machine-generated data to train the model. |
| Outcome: | The proposed model can be trained on a massive hate speech dataset with positive samples . it can be generalized and reduce identity term bias, the authors show . |
Marking Code Without Breaking It: Code Watermarking for Detecting LLM-Generated Code (2026.findings-eacl)
Copied to clipboard
| Challenge: | Existing methods for identifying LLM-generated code are limited by syntax-critical tokens, which can introduce syntax errors. |
| Approach: | They propose a syntax-aware watermarking method that embeds watermarks only in non-syntactic tokens and preserves code integrity. |
| Outcome: | The proposed method outperforms baseline methods on Python, C++, and Java. |