Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks86 from public data
- Post-Training Quantization on Diffusion Models376
This work accelerates generation of diffusion model (DM) generation from the perspective of compressing the noise estimation network and devise a DM-specific PTQ method, which especially targets the unique multi-time-step structure of DMs.
- LLM Inference Unveiled: Survey and Roofline Model Insights232
This survey systematically collates the latest advancements in efficient LLM inference, covering crucial areas such as model compression, algorithm improvements, and both hardware and system-level enhancements, and introduces a framework based on roofline model for systematic analysis of LLM inference techniques.
- RPTQ: Reorder-based Post-training Quantization for Large Language Models148
This paper identifies that the challenge in quantizing activations in LLMs arises from varying ranges across channels, rather than solely the presence of outliers, and introduces a quantization method called RPTQ, which utilizes a reorder-based approach.
- Vertically Federated Graph Neural Network for Privacy-Preserving Node Classification132
This paper proposes VFGNN, a federated GNN learning paradigm for privacy-preserving node classification task under data vertically partitioned setting, which can be generalized to existing GNN models.
- Parameter-Efficient Fine-Tuning with Discrete Fourier Transform91
This work introduces FourierFT, which treats $\Delta W$ as a matrix in the spatial domain and learns only a small fraction of its spectral coefficients, and implements the inverse discrete Fourier transform to recover $\Delta W$.
- HASCO: Towards Agile HArdware and Software CO-design for Tensor Computation89
This work proposes an agile co-design approach HASCO that provides an efficient HW/SW solution to dense tensor computation and develops a multi-objective Bayesian optimization algorithm to explore hardware optimization.
- ASFGNN: Automated separated-federated graph neural network89
An Automated Separated-Federated Graph Neural Network (ASFGNN) learning paradigm is proposed, which decouples the training of GNN into two parts: the message passing part that is done by clients separately, and the loss computing part that was learnt by clients federally.
- DrugOOD: Out-of-Distribution Dataset Curator and Benchmark for AI-Aided Drug Discovery – a Focus on Affinity Prediction Problems with Noise Annotations81
DrugOOD offers an automated dataset curator with user-friendly customization scripts, rich domain annotations aligned with biochemistry knowledge, realistic noise level annotations, and rigorous benchmarking of SOTA OOD algorithms, as opposed to only providing fixed datasets.
- Practical Privacy Preserving POI Recommendation78
This article proposes a novel Privacy preserving POI Recommendation (PriRec) framework that keeps users’ private raw data and models in users' own hands, and protects user privacy to a large extent, and applies PriRec in real-world datasets, and comprehensive experiments demonstrate that it achieves comparable or even better recommendation accuracy.
- Uncertainty in Natural Language Processing: Sources, Quantification, and Applications67
This survey provides a comprehensive review of uncertainty-relevant works in the NLP field and categorizes the sources of uncertainty in natural language into three types, including input, system, and output.
- Calibrating Multimodal Learning52
This paper finds that the confidence estimated by current models could even increase when some modalities are corrupted, and proposes a novel regularization technique, i.e., Calibrating Multimodal Learning (CML) regularization, to calibrate the predictive confidence of previous methods.
- P3SGD: Patient Privacy Preserving SGD for Regularizing Deep CNNs in Pathological Image Classification45
A novel stochastic gradient descent scheme, named patient privacy preserving SGD (P3SGD), which performs the model update of the SGD in the patient level via a large-step update built upon each patient's data to protect privacy and regularize the CNN model.
- Secure Social Recommendation Based on Secret Sharing44
A SEcure SOcial RECommendation (SeSoRec) framework which can collaboratively mine knowledge from social platform to improve the recommendation performance of the rating platform, and securely keep the raw data of both platforms is proposed and its correctness and security theoretically is proved.
- Adapting Large Language Models for Content Moderation: Pitfalls in Data Engineering and Supervised Fine-tuning43
This paper presents a complete process, from data collection and construction to model training and overfitting elimination, for fine-tuning LLMs in vertical domain deployments, and reveals that incorporating reasoning processes during the fine-tuning of LLMs can effectively alleviate overfitting.
- 39
- 39
- Privacy-Preserving Graph Neural Network for Node Classification.39
This paper proposes a Privacy-Preserving GNN (PPGNN) learning paradigm for node classification task, which can be generalized to existing GNN models and demonstrates that PPGNN significantly outperforms theGNN models trained on the isolated data and has comparable performance with the traditional GNNtrained on the mixed plaintext data.
- 39
- A Survey of Trustworthy Graph Learning: Reliability, Explainability, and Privacy Protection33
This survey provides a comprehensive review of recent leading approaches in the TwGL field from three dimensions, namely, reliability, explainability, and privacy protection, and gives a general categorization for existing work and review typical work for each category.
- S2DNAS: Transforming Static CNN Model for Dynamic Inference via Neural Architecture Search33
This paper introduces a general framework, S2DNAS, which can transform various static CNN models to support dynamic inference via neural architecture search and proposes a reinforcement learning based approach to automatically search for the optimal CNN architecture in the generated space.
- 31
- NAS4RRAM: neural network architecture search for inference on RRAM-based accelerators31
This work takes the idea of network architecture search (NAS) to design networks with high prediction accuracy that meet the requirements of the RRAM-based accelerators, and proposes a framework called NAS4RRAM to search for the optimal network on the given R RAM-based accelerator.
- 31
- Reducing Overfitting in Deep Convolutional Neural Networks Using Redundancy Regularizer27
A novel regularizer is proposed to reduce kernel redundancy in a deep CNN model and prevent overfitting and results demonstrate that kernel redundancy is significantly removed and overfitting is substantially reduced with even better performance achieved.
- NTPP: Generative Speech Language Modeling for Dual-Channel Spoken Dialogue via Next-Token-Pair Prediction26
This work systematically explore the use of dual-channel speech data in the context of modern large language models, and introduces a novel generative modeling paradigm, Next-Token-Pair Prediction (NTPP), to enable speaker-independent dual-channel spoken dialogue learning using decoder-only architectures for the first time.
- 26
- Characterizing Membership Privacy in Stochastic Gradient Langevin Dynamics25
It is suggested that the SGLD algorithm can not only reduce the information leakage but also improve the generalization ability of the DNN models in real-world applications.
- Rethinking and Simplifying Bootstrapped Graph Latents24
SGCL is presented, a simple yet effective GCL framework that utilizes the outputs from two consecutive iterations as positive pairs, eliminating the negative samples and can achieve competitive performance with fewer parameters, lower time and space costs, and significant convergence speedup.
- Trustworthy Graph Learning24
This tutorial mainly covers the key achievements of trustworthy graph learning in recent years and discusses three essential topics, that is, the reliability of DGL against inherent noise, distribution shift and adversarial attack, explainability methods, and privacy protection for DGL.
- 23
- A Review and Experimental Evaluation on Split Learning21
This work implements several typical SL paradigms and performs extensive experiments to compare their performance in different scenarios on four widely used datasets, and provides extensive engineering advice and research insights for SL paradigms.
- Variate Associated Domain Adaptation for Unsupervised Multivariate Time Series Anomaly Detection21
A Variate Associated Domain Adaptation Method Combined with a Graph Deviation Network (VANDA) for MTS-AD is proposed, which outperforms the benchmark methods, and significantly improves the AD performance of the target domain by effectively utilizing the source domain knowledge.
- Learning With Noisy Labels Over Imbalanced Subpopulations18
A novel LNL method to deal with noisy labels and imbalanced subpopulations simultaneously that first leverages sample correlation to estimate samples’ clean probabilities for label correction and then utilizes corrected labels for distributionally robust optimization (DRO) to further improve the robustness.
- 18
- 18
- 18
- Step-On-Feet Tuning: Scaling Self-Alignment of LLMs via Bootstrapping17
This pioneering exploration delves into the impact of bootstrapping self-alignment on large language models and reveals that bootstrapping self-alignment markedly surpasses the single-round approach, by guaranteeing data diversity from in-context learning.
- Benchmarking the Reliability of Post-training Quantization: a Particular Focus on Worst-case Performance13
This paper investigates the influence of calibration set distribution variations, calibration paradigm selection, and data augmentation or sampling strategies on PTQ reliability on various commonly-used PTQ methods.
- SAILOR: Structural Augmentation Based Tail Node Representation Learning12
This work explores how the deficiency of structural information deteriorates the performance of tail nodes and proposes a general structural augmentation based tailno de representation learning framework, dubbed as øurs, which can jointly learn to augment the graph structure and extract more informative representations for tail nodes.
- 12
- Probing the Safety Response Boundary of Large Language Models via Unsafe Decoding Path Generation11
It is argued that: even when an LLM appears to successfully block harmful queries, there may still be hidden vulnerabilities that could act as ticking time bombs and be exploited via prompt optimization such as soft prompts on images.
- Privacy Matters: Vertical Federated Linear Contextual Bandits for Privacy Protected Recommendation10
A customized encryption scheme named orthogonal matrix-based mask mechanism (O3M), a tailored component for contextual bandits by carefully exploiting their shared structure, can ensure privacy protection while avoiding expensive conventional cryptographic techniques.
- ImDrug: A Benchmark for Deep Imbalanced Learning in AI-aided Drug Discovery10
ImDrug is introduced, a comprehensive benchmark with an open-source Python library which consists of 4 imbalance settings, 11 AI-ready datasets, 54 learning tasks and 16 baseline algorithms tailored for imbalanced learning that demonstrates that the existing algorithms fall short of solving medicinal and pharmaceutical challenges in the data imbalance scenario.
- Toward Scalable and Privacy-preserving Deep Neural Network via Algorithmic-Cryptographic Co-design10
This article proposes SPNN—a Scalable and Privacy-preserving deep Neural Network learning framework, from an algorithmic-cryptographic co-perspective, and proposes using two types of cryptographic techniques, i.e., secret sharing and homomorphic encryption, for the isolated data holders to conduct private-data-related computations privately and cooperatively.
- Measuring Diversity in Synthetic Datasets9
DCScore is introduced, a novel method for measuring synthetic dataset diversity from a classification perspective that formulates diversity evaluation as a sample classification task, leveraging mutual relationships among samples.
- EARBench: Towards Evaluating Physical Risk Awareness for Task Planning of Foundation Model-based Embodied AI Agents9
RiskAwareBench is proposed, an automated framework designed to assess physical risks awareness in LLM-based embodied agents, and compile the PhysicalRisk dataset, encompassing diverse scenarios with associated safety tips, observations, and instructions.
- VDC: Versatile Data Cleanser based on Visual-Linguistic Inconsistency by Multimodal Large Language Models9
This paper proposes versatile data cleanser (VDC) leveraging the surpassing capabilities of multimodal large language models (MLLM) in cross-modal alignment and reasoning to capture the semantic inconsistency between modalities.
- SLPerf: a Unified Framework for Benchmarking Split Learning9
This work proposes SLPerf, a unified research framework and open research library for SL, and conducts extensive experiments on four widely-used datasets under both IID and Non-IID data settings to help facilitate SL algorithm development and fair performance comparisons.
- 8
- Federated Nearest Neighbor Machine Translation7
A novel federated nearest neighbor (FedNN) machine translation framework that leverages one-round memorization-based interaction to share knowledge across different clients to build low-overhead privacy-preserving systems.
- 7
- G2C: A Generator-to-Classifier Framework Integrating Multi-Stained Visual Cues for Pathological Glomerulus Classification7
The approach, named generator-to-classifier (G2C), is a two-stage framework that combines visual cues from different stains for prediction, and boosts classification performance, suggesting more effective visual cues are extracted in an automatic way.
- Spurious Feature Eraser: Stabilizing Test-Time Adaptation for Vision-Language Foundation Model6
This work introduces a test-time prompt tuning paradigm that optimizes a learnable prompt, thereby compelling the model to exploit invariant features while disregarding decision shortcuts during the inference phase, and conducts comparative analysis of the proposed method against various approaches which validates the significant superiority.
- Language Agents for Detecting Implicit Stereotypes in Text-to-image Models at Scale6
A novel agent architecture tailored for stereotype detection in text-to-image models is introduced, capable of accommodating free-form detection tasks and can autonomously invoke various tools to facilitate the entire process, from generating corresponding instructions and images, to detecting stereotypes.
- 6
- A Label Disambiguation-Based Multimodal Massive Multiple Instance Learning Approach for Immune Repertoire Classification5
A novel label disambiguation-based multimodal massive multiple instance learning approach (LaDM³IL) for immune repertoire classification that adapts the instance-level MIL paradigm to deal with the issue of high computational cost and employs a specially-designed label disambiguation module for label correction, mitigating the impact of misleading supervision.
- DRFLM: Distributionally Robust Federated Learning with Inter-client Noise via Local Mixup5
A general framework to solve the above two challenges simultaneously, using distributionally robust optimization to mitigate the negative effects caused by data heterogeneity paradigm to sample clients based on a learnable distribution at each iteration is proposed.
- Nebula5
A privacy-preserving collaborative learning system in Ant Financial, named Nebula, which is built upon a ring-allreduce MPI based distributed framework and able to scale up to tens of millions of data samples with hundreds of thousands of features and achieve more than 100x speedup compared with the existing state-of-the-art implementations.
- IBCircuit: Towards Holistic Circuit Discovery with Information Bottleneck4
IBCircuit is an optimization framework for holistic circuit discovery that can be applied to any given task without tediously corrupted activation design and identifies more faithful and minimal circuits in terms of critical node components and edge components compared to recent related work.
- Enabling Early Identification of Malignant Vertebral Compression Fractures Through 2.5D Convolutional Neural Network Model With CT Image Analysis4
A 2.5D convolutional neural network model leveraging CT imaging to facilitate the early detection of malignant vertebral compression fractures (MVCFs) and its performance in identifying MVCF patients was significantly superior to that of the 2D and 3D CNN models.
- Reweighted Mixup for Subpopulation Shift4
This work proposes a simple yet practical framework, called reweighted mixup (RMIX), to mitigate the overfitting issue in over-parameterized models by conducting importance weighting on the ''mixed'' samples by leveraging reweighting in mixup.
- COPPER: a combinatorial optimization problem solver with processing-in-memory architecture4
A specialized processing-in-memory hardware architecture named COPPER based on the memristor is designed, capable of efficiently running the modified quantized CSA algorithm and supporting the pipeline further acceleration, and shows that COPPER can perform CSA remarkably well in both speed and energy.
- 4
- 4
- AuditAgent: Expert-Guided Multi-Agent Reasoning for Cross-Document Fraudulent Evidence Discovery3
This work introduces a novel multi-agent reasoning framework AuditAgent, enhanced with auditing domain expertise, for fine-grained evidence chain localization in financial fraud cases, and demonstrates that this method substantially outperforms General-Purpose Agent paradigm in both recall and interpretability.
- 3
- 3
- A Versatile Deep Graph Contrastive Learning Framework for Single-cell Proteomics Embedding3
A novel versatile framework scPROTEIN, composed of peptide uncertainty estimation based on a multi-task heteroscedastic regression model and cell embedding learning based on graph contrastive learning designed for single-cell proteomic data analysis, is proposed.
- Attention Paper: How Generative AI Reshapes Digital Shadow Industry?2
A technical analysis of the challenges and opportunities of AIGC from upstream, midstream, and downstream paths of black/shadow industries and suggest future directions for improving existing risk control systems are provided.
- 1
- 1
- Sparsification and Reconstruction from the Perspective of Representation Geometry1
The SAEMA is proposed, which validates the stratified structure of the representation by observing the variability of the rank of the symmetric semipositive definite (SSPD) matrix corresponding to the modal tensor unfolded along the latent tensor with the level of noise added to the residual stream.
- Vertical Federated Linear Contextual Bandits1
The proposed protocols can perfectly recover the service quality of centralized bandit algorithms while achieving a satisfactory runtime efficiency, which is theoretically proved and analyzed in this paper.
- Automatic Knowledge Fusion in Transferrable Networks for Semantic Text Matching1
This work proposes a novel method to automatically fuse transferable network architecture that has better performance than the state-of-the-art transfer learning architectures.
- 1
- 1
- 1
- –
- –
- –
- ICAT: Incident-Case-Grounded Adaptive Testing for Physical-Risk Prediction in Embodied World Models–
This work proposes ICAT, which grounds testing in real incident reports and safety manuals by building structured risk memories and retrieving/composing them to constrain the generation of risk cases with causal chains and severity labels.
- –
- –
- Is GPT4 a Good Trader?–
This study aims to examine the fidelity of GPT-4's comprehension of classic trading theories and its proficiency in applying its code interpreter abilities to real-world trading data analysis, and could pave the way for a synergistic amalgamation of human expertise and AI-driven insights in the realm of trading.
- Semantic Equivariant Mixup–
Different from previous mixup variants, which tend to over-focus on the label-related information, the proposed method aims to preserve richer semantic information in the input with semantic-equivariance assumption, thereby improving the robustness of the model against distribution shifts.
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.