Institution profile

Illinois Institute of Technology

Academic institutionnorthamerica · us
Official website
Research library190linked papers
Opportunities0open roles
Selected work

Representative Papers

Authorship Attribution in the Era of LLMs: Problems, Methodologies, and Challenges

Aug 16, 2024arXiv.org

The rise of large language models (LLMs) has intensified authorship attribution challenges—namely, distinguishing human-authored text from LLM-generated content and resolving ambiguous attribution in human-AI collaborative writing. To address this, we propose the first four-category authorship taxonomy for the LLM era: human-authored, LLM-generated, LLM-attributed, and human-AI collaborative. We systematically survey detection methodologies across four paradigms: statistical features, neural representations, attribution graphs, and prompt engineering—covering models including BERT, RoBERTa, and Llama, as well as watermarking and probability calibration techniques. Further, we introduce a unified evaluation framework balancing cross-domain generalizability and decision interpretability, and establish the field’s first dynamically updated resource repository (llm-authorship.github.io). Our work provides both theoretical foundations and a practical roadmap for enhancing detection accuracy and transparency in LLM-era authorship attribution.

6 citationsRead paper

Causality-Aware Temporal Projection for Video Understanding in Video-LLMs

Jan 05, 2026arXiv.org

This work addresses the challenge that existing video large language models often fail to preserve strict temporal order and causal consistency due to bidirectional temporal modeling, which disrupts chronological sequence. To resolve this, the authors propose the V-CORE framework, which incorporates Learnable Spatial Aggregation (LSA) to retain spatial interactions and introduces a Causal-Aware Temporal Projector (CATP) to enforce unidirectional temporal information flow, thereby ensuring causal consistency. The approach explicitly models temporal ordering constraints through block-wise causal attention and dynamic causal sink tokens. Furthermore, it adopts 4-bit QLoRA fine-tuning with a frozen LLM backbone for computational efficiency. Evaluated on NExT-QA, the method achieves 61.2% accuracy, with notable improvements of 3.5% and 5.2% on temporal and causal reasoning subsets, respectively, demonstrating the effectiveness of the proposed temporal constraint mechanism.

1 citationsRead paper

Conditional Neural ODE for Longitudinal Parkinson's Disease Progression Forecasting

Nov 06, 2025

Parkinson’s disease (PD) exhibits highly heterogeneous and irregular longitudinal brain morphological changes, posing challenges for existing RNN- or Transformer-based longitudinal modeling approaches—particularly in handling sparse, irregularly sampled MRI data and capturing inter-individual variability in disease onset timing and progression rates. To address these limitations, we propose CNODE, a continuous-time framework grounded in neural ordinary differential equations (Neural ODEs) to model smooth, interpretable brain structural dynamics. CNODE incorporates a conditional encoding mechanism to accommodate irregular sampling intervals and jointly learns patient-specific disease onset times and progression velocities. By aligning individual trajectories onto a shared pathological progression manifold, it enables personalized “digital twin” forecasting. Evaluated on the PPMI dataset, CNODE achieves statistically significant improvements over state-of-the-art methods, especially in long-term trajectory prediction accuracy.

1 citationsRead paper

The Power of Graph Signal Processing for Chip Placement Acceleration

Feb 24, 2025

VLSI placement faces challenges of high computational complexity and slow convergence in nonlinear optimization. Existing GCN-based deep learning methods rely heavily on large-scale labeled data and prolonged training, limiting generalizability. This paper proposes GiFt, a training-free, parameter-free graph signal processing framework that leverages circuit netlist topology as structural prior. GiFt employs multi-resolution graph filtering to efficiently smooth and model placement signals, thereby accelerating the convergence of analytical placers without supervision. Crucially, GiFt eliminates the need for any training or hyperparameter tuning, drastically lowering deployment overhead. Experiments demonstrate that GiFt reduces total runtime by over 45% compared to the GPU-accelerated state-of-the-art placer DREAMPlace, while maintaining or even improving placement quality—achieving, for the first time, training-free, high-efficiency, high-quality, graph-structure-driven placement optimization. Both algorithmic design and empirical performance set new state-of-the-art benchmarks.

1 citationsRead paper
Recent publications

Latest Papers

Conformity Breaks Conformal Prediction

Sep 03, 2026

研究解决了多智能体系统中一致性预测失效的问题,通过分析模型在同伴压力下的评分机制变化,并揭示了标准校正方法对此问题无效的原因。

0 citationsRead paper