Institution profile

Purdue University

Academic institutionnorthamerica · us
Official website
Research library1,868linked papers
Opportunities0open roles
Selected work

Representative Papers

Exponential Lower Bounds for Locally Decodable and Correctable Codes for Insertions and Deletions

Nov 01, 2021IEEE Annual Symposium on Foundations of Computer Science

This work investigates the existence of locally decodable codes (LDCs) under insertion-deletion (insdel) errors. Addressing a long-standing open conjecture, we prove—*for the first time*—that no 2-query linear insdel LDC exists. Moreover, for any constant query complexity $q geq 3$, we establish an exponential lower bound $exp(Omega(n))$ on the code length, significantly stronger than the polynomial bounds known for Hamming-error LDCs. Our approach constructs a hard insdel error distribution and combines information-theoretic analysis with novel coding reduction techniques. This reveals a fundamental separation between insdel LDCs and Hamming LDCs—a separation that persists even in the adaptive decoding and private-key settings. The results characterize the theoretical limits of local error correction against synchronization errors and provide the first tight lower bounds for insdel coding.

7 citations2 influentialRead paper

Jailbreaking and Mitigation of Vulnerabilities in Large Language Models

Oct 20, 2024arXiv.org

Existing research on large language models (LLMs) lacks a unified taxonomy for prompt injection and jailbreaking attacks, and insufficiently evaluates defenses under dynamic, interactive scenarios. Method: We propose the first four-dimensional attack taxonomy—spanning prompt-level, model-level, multimodal, and multilingual pathways—and develop a robust alignment framework tailored to interactive settings, alongside a novel automated jailbreaking detection method. We further conduct systematic defense benchmarking, bias diagnosis of existing evaluation benchmarks, and multi-dimensional security measurement. Contribution/Results: Our analysis reveals critical failure modes of current defenses in dynamic interactions, identifies key research gaps—including ethical implications and data bias—and delivers the first comprehensive technical roadmap for LLM safety alignment.

7 citationsRead paper

Emotion Recognition Using Convolutional Neural Networks

Jan 13, 2019IMAWM

Real-time, fine-grained facial emotion recognition (FER) remains challenging due to the need for joint classification of discrete emotion categories and continuous intensity estimation across heterogeneous input modalities (static images and video streams). Method: We propose an end-to-end FER system supporting both static and real-time video inputs, trained to recognize seven basic emotions (anger, disgust, fear, happiness, neutral, sadness, surprise) via simultaneous coarse-grained classification and continuous intensity regression. Our approach introduces a custom full-pipeline framework—including controlled data acquisition, grayscale normalization, tailored data augmentation, and a lightweight CNN architecture—optimized jointly on a proprietary dataset and two public benchmarks (CK+ and RAF-DB). Contribution/Results: To our knowledge, this is the first work achieving concurrent high-accuracy classification (mean accuracy >80%) and interpretable intensity regression within a single model. With <50 ms inference latency per frame, the system satisfies real-time video processing requirements, demonstrating the feasibility and robustness of CNNs for industrial-grade emotion sensing.

6 citationsRead paper

Towards Sustainable Large Language Model Serving

Dec 31, 2024

Prior studies on LLM carbon footprint overlook embodied emissions (e.g., chip manufacturing) and regional grid heterogeneity, leading to incomplete sustainability assessments. Method: We propose the first holistic sustainability evaluation framework jointly modeling operational carbon emissions (during inference) and embodied emissions (from GPU fabrication), enabling cross-hardware (RTX6000 Ada/T4) and cross-regional (varying grid carbon intensities) analysis. Using empirical measurements of LLaMA-1B/3B/7B inference power consumption, die area, and memory parameters, we quantify embodied emissions and integrate region-specific grid carbon intensities for end-to-end emission accounting. Contribution/Results: Experiments reveal strong coupling among hardware generation, model scale, and grid cleanliness in determining total carbon footprint. We open-source a reproducible assessment toolkit, providing theoretical foundations and actionable optimization pathways for low-carbon LLM deployment.

4 citationsRead paper

On Memorization and Privacy risks of Sharpness Aware Minimization

Sep 30, 2023arXiv.org

This work investigates whether Sharpness-Aware Minimization (SAM), while improving generalization, exacerbates membership privacy risks. We find that SAM significantly enhances memorization of anomalous samples—increasing memorization rate by 12.3% over SGD—and raises membership inference attack success to 89.7%, indicating its generalization gain stems from reinforced fitting of memorizable instances. To address this, we propose a novel memorization metric and provide the first systematic evidence that SAM inherently entails a heightened privacy–accuracy trade-off. We further design a joint gradient regularization and noise injection mechanism that reduces privacy risk by 41% while incurring less than 0.8% accuracy degradation. Our findings offer new insights into the intrinsic privacy implications of optimization algorithms and deliver a practical, privacy-aware training framework.

3 citationsRead paper
Recent publications

Latest Papers