Yang Liu

Introducing a technique for better predictions and confidence intervals in decision-making.

A new method improves detection of changing objects using combined knowledge sources.

This paper presents new sampling methods for better generative model performance.

A new method to protect privacy in text-to-image models.

Explore the importance of reliability in assessing psychological traits and behaviors.

Scientists improve detection methods for dark photons, potential candidates for dark matter.

A look into phases of universe expansion and string theory's role.

New methods improve detection of Ponzi schemes in cryptocurrency investments.

This article discusses fairness measurement in recommendation systems and proposes sampling methods.

Examining how fairness in machine learning can evolve across decisions and time.

UGradSL offers a practical solution for removing sensitive information from machine learning models.

This article discusses the challenges of machine unlearning and a new approach to balance privacy and accuracy.

Samba efficiently manages long sequences for better language processing.

A new detector aims to probe unexplored areas of particle physics.

Examining the impact of deprecated APIs on LLM code suggestions.

A look at how randomized quasi-Monte Carlo methods estimate integrals under various conditions.

Introducing a framework to better assess common factor models using generalized residuals.

FuseGen combines multiple models for better quality synthetic data in machine learning.

New framework improves modeling of complex physical systems using discrete symmetry.

This paper explores fairness and stability in machine learning models affected by their predictions.

A look into topological properties and dynamics in quantum systems.

Recent findings question our understanding of the universe's expansion and structure formation.

SimXRD offers millions of simulated XRD patterns for improved crystal classification.

A two-stage framework improves Python's type inference, enhancing code quality and efficiency.

Rethinking how machines interact for more natural conversations.

TongGu simplifies the understanding of Classical Chinese with specialized techniques.

A new method enhances SiN circuits for better performance and scalability.

A reliable earbud-based system monitors breathing rates during various daily activities.

New IRIS technique enhances CEST MRI image quality for better diagnostics.

A new method improves the search for mathematical expressions from data.

VQA-Diff combines techniques to enhance 3D vehicle modeling from real-world images.

A new dataset enhances machine learning in understanding 3D environments and language.

Fisher zeros reveal insights into phase transitions in quantum mechanics.

Researchers develop split ribozyme biosensors to track RNA in living plants.

A factory in Huizhou aims to deepen our knowledge of particles and the universe.

New methods improve black-box attacks using unlabeled data for machine learning.

New AI tools are simplifying music editing with innovative techniques and improved precision.

A new framework enhances knowledge graph completion efficiency and accuracy using large language models.

A new model enhances battery range predictions for electric bikes.

A guide to understanding reliability in psychology assessments.

This study presents a method to enhance tracing of security patches.

SuperVINS enhances robot navigation with deep learning techniques for improved mapping.

Movable antennas adapt to improve communication performance in various environments.

A method to identify weaknesses in language models for better performance.

Techniques to safeguard personal images from misuse by generative models.

Examining the balance between geolocation technology and user privacy.

Vulseye improves smart contract safety through targeted fuzz testing and analysis.

Introducing DUCD, a privacy-focused method to protect DNNs from adversarial attacks.

A new method improves detection of harmful prompts in language models.

ADC streamlines dataset creation, improving speed and accuracy in machine learning.

A new approach for generating controlled financial market scenarios.

A look at how Knowledge Distillation enhances recommendation systems' speed and efficiency.

This method enhances the learning process of knowledge graphs through structured training.

A new method locates video events using large pre-trained models without specific training.

A look into the methods and applications of facial expression recognition technology.

ResVG model improves object recognition in images through enhanced detail and relationships.

S-money tokens offer a new way to conduct secure financial transactions with privacy.

A new method enables language models to correct their own mistakes in math.

A framework using memory tokens improves video understanding and interaction.

MarS leverages generative models to simulate realistic financial market scenarios.

Assessing large language models' ability to handle privacy regulations.

A look into cosmic birefringence and its significance in astrophysics.

Learn about the remarkable structure and function of elephant trunks.

New methods improve decision-making in complex resource allocation problems.

LogicPro enhances logical reasoning in AI using algorithm problems and coded examples.

Enhancing detection methods for harmful packages in software repositories.

New methods enhance the accuracy of influence functions in large models.

As technology advances, software security faces new challenges and opportunities.

A look at how the Sun's magnetic fields affect space weather and technology on Earth.

Exploring the link between personality traits and code generation effectiveness using LLMs.

This article discusses how models improve their reasoning through self-training and learning from mistakes.

A new framework aims to uncover biases in role-playing scenarios of language models.

Discover how MAP Protocol is bridging blockchain gaps for seamless transactions.

NeuroFly automates neuron reconstruction, saving time for researchers and enhancing brain studies.

A look at how LLMs process language through reasoning techniques.

Researchers achieve long-distance secure communication with new technology.

A database to combat backdoor defects in deep learning models.

Learn how text watermarking secures your content without altering its meaning.

Scientists aim to uncover the mysteries of dark matter through dark scalar particles.

A new platform where robots can learn interaction and skills like humans.