Papers by Viet Pham
PARASITE: Conditional System Prompt Poisoning to Hijack LLMs (2026.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) are increasingly deployed via third-party system prompts downloaded from public marketplaces. |
| Approach: | They propose a framework that optimizes system prompts to trigger LLMs to output compromised responses only for specific queries. |
| Outcome: | The proposed framework achieves up to 70% F1 reduction on targeted queries with minimal degradation to general capabilities. |
The Dangers of Indirect Prompt Injection Attacks on LLM-based Autonomous Web Navigation Agents: A Demonstration (2025.emnlp-demos)
Copied to clipboard
| Challenge: | Large Language Model (LLM)-integrated applications are becoming more popular to support, augment, and automate tasks. |
| Approach: | They propose to embed universal adversarial triggers in webpage HTML to hijack agents . they also use a browser-gym agent powered by Llama-3.1 to test their system . |
| Outcome: | The proposed system software is released under the MIT License . |
Lifelong Event Detection via Optimal Transport (2024.emnlp-main)
Copied to clipboard
| Challenge: | Continual event detection (CED) is a challenging task due to catastrophic forgetting, where learning new tasks hampers performance on previous ones. |
| Approach: | They propose a method that leverages optimal transport principles to align the optimization of a classification module with the intrinsic nature of each class, as defined by their pre-trained language modeling. |
| Outcome: | The proposed method outperforms state-of-the-art methods on MAVEN and ACE datasets and is a pioneering solution in continual event detection. |
VN-MTEB: Vietnamese Massive Text Embedding Benchmark (2026.findings-eacl)
Copied to clipboard
| Challenge: | a lack of large-scale test datasets makes it difficult to evaluate AI models before deploying them in real-world projects. |
| Approach: | They propose a Vietnamese benchmark for embedding models that leverages large language models and embeddable models to translate and filter samples from the Massive Multilingual Text Embedding Benchmark. |
| Outcome: | The proposed benchmark outperforms existing models in Vietnamese and English tasks with 41 datasets. |