Papers by Huixuan Zhang
Image Matters: A New Dataset and Empirical Study for Multimodal Hyperbole Detection (2024.lrec-main)
Copied to clipboard
| Challenge: | linguistic detection of hyperbole is an important part of understanding human expression . studies on hyperbolic expressions focus on text modality, but social media can be used to detect it . |
| Approach: | They propose to use a multimodal detection dataset to study hyperbole detection . they treat text and image as two modalities and evaluate pre-trained encoders . |
| Outcome: | The proposed dataset is constructed from five different keywords and shows its performance. |
MC-MKE: A Fine-Grained Multimodal Knowledge Editing Benchmark Emphasizing Modality Consistency (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing benchmarks for knowledge editing in multimodal large language models focus on limited scenarios due to the lack of rigorous definition of multimodal knowledge. |
| Approach: | They propose a decomposed definition of multimodal knowledge and a benchmark to evaluate it. |
| Outcome: | The proposed method reveals that it is difficult to define multimodal knowledge editing in LLMs. |
Minos: A Multimodal Evaluation Model for Bidirectional Generation Between Image and Text (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing evaluation models struggle to achieve consistent performance across image-to-text (I2T) and text-to image (T2I) tasks. |
| Approach: | They construct a multimodal evaluation model using a large multimodal dataset and rigorous quality control strategies to train it. |
| Outcome: | The proposed model achieves state-of-the-art evaluation performance across 16 out-of domain datasets covering both I2T and T2I tasks among all open-source multimodal evaluation models and remain competitive with closed-source models. |
ICR Probe: Tracking Hidden State Dynamics for Reliable Hallucination Detection in LLMs (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for hallucination detection rely on static and isolated representations, overlooking their dynamic evolution across layers. |
| Approach: | They propose a method which captures the cross-layer evolution of hidden states and propose 'ICR Probe' which capture the evolution of the hidden states. |
| Outcome: | The proposed method achieves superior performance with significantly fewer parameters and ablation studies offer deeper insights into the underlying mechanism of the method, improving its interpretability. |
JointCQ: Improving Factual Hallucination Detection with Joint Claim and Query Generation (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods for detecting factual hallucinations in generated content exhibit limitations in the first two stages of the halluciation detection pipeline. |
| Approach: | They propose a joint claim-and-query generation framework that can detect factual hallucinations in generated content. |
| Outcome: | The proposed method outperforms existing methods on open-domain QA hallucination detection benchmarks. |
Exploring and Evaluating Multimodal Knowledge Reasoning Consistency of Multimodal Large Language Models (2025.findings-emnlp)
Copied to clipboard
| Challenge: | MLLMs have achieved significant breakthroughs in understanding across text and vision, but current models still face inconsistencies in reasoning outcomes. |
| Approach: | They propose to evaluate multimodal large language models using a multimodal knowledge reasoning dataset to examine the extent of consistency degradation. |
| Outcome: | The proposed evaluation tasks show that MLLMs are inefficient at integrating knowledge across modalities . |
DecoCal: Decoding with Calibration in Diffusion Large Language Models (2026.acl-long)
Copied to clipboard
| Challenge: | Diffusion Large Language Models (DLLMs) generate text via iterative token denoising . but decoding is challenging, with many tokens appearing predictable early . |
| Approach: | They propose a Decoding framework that performs Calibration of token-level confidence across diffusion steps and leverages the calibrated results to guide decoding decisions. |
| Outcome: | Experiments on multiple DLLMs and benchmarks show that DecoCal improves generation accuracy compared to existing strategies. |
PaCoST: Paired Confidence Significance Testing for Benchmark Contamination Detection in Large Language Models (2024.findings-emnlp)
Copied to clipboard
| Challenge: | Large language models are trained on vast amounts of data, which may unintentionally or intentionally include data from commonly used benchmarks. |
| Approach: | They propose a set of requirements that practical contamination detection methods should follow to effectively detect benchmark contamination in large language models. |
| Outcome: | The proposed method detects whether the model is significantly more confident under the original benchmark. |
DAMON: A Dialogue-Aware MCTS Framework for Jailbreaking Large Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for multi-turn attacks mainly utilize a predefined dialogue pattern, limiting their effectiveness in realistic situations. |
| Approach: | They propose a multi-turn jailbreak attack method that leverages Monte Carlo Tree Search to explore multi-turned conversational spaces and identifies sub-instruction sequences that induce harmful responses. |
| Outcome: | The proposed method can induce undesired behaviors across five LLMs and three datasets. |
R-Bind: Unified Enhancement of Attribute and Relation Binding in Text-to-Image Diffusion Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing approaches to improve semantic binding require costly retraining or focus on only correctly generating attributes of entities. Existing methods focus on correctly generating attributes, ignoring the cruciality of correctly forming relations between entities. |
| Approach: | They propose a training-free method that improves both entity-attribute and entity-relation-entity binding by introducing three inference-time optimization losses that adjust attention maps during generation. |
| Outcome: | The proposed method improves both entity-attribute and entity-relation-entity binding without additional training. |
Tracing Training Footprints: A Calibration Approach for Membership Inference Attacks Against Multimodal Large Language Models (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Existing methods to improve difficulty calibration for Multimodal Large Language Models only consider text input . visual embeddings in training data reduce effectiveness of these methods . |
| Approach: | They propose a method to detect member samples in poorly generalized local manifolds by visual embeddings. |
| Outcome: | The proposed method surpasses existing methods. |