Jie Wang

Full Professor@University of Science and Technology of China·中国·OpenReview

研究方向

deep learning · graph neural networks · reinforcement learning · large language models · learn2optimize · AI4EDA · AI4Science · knowledge graph · deep reinforcement learning · natural language processing · machine learning · sparse learning · large scale optimization · SVM · classification · sparse models · image processing · image segmentation

Jie Wang

Evolving Graph Structured Programs for Circuit Generation with Large Language Models

A Hierarchical Circuit Symbolic Discovery Framework for Efficient Logic Optimization

Following the Navigation: Enhancing Small Language Models Contextual Reasoning with LLM Guidance

Boosting Multi-Domain Reasoning of LLMs via Curvature-Guided Policy Optimization

Latent-Guided Reasoning: Empowering Small LLMs with Large-Model Thinking

Accelerating Eigenvalue Dataset Generation via Chebyshev Subspace Filter

Preference Optimization via Key-step Error Exploration for Multi-step Reasoning in LLMs

Tackling Heavy-Tailed Q-Value Bias in Offline-to-Online Reinforcement Learning with Laplace-Robust Modeling

OptiVer: Unleashing the Power of LLMs for Optimization Modeling via Dual-Side Verification

Learning A Linear Delay Surrogate Model for Timing-Driven Chip Global Placement

CAPO: Conflict-Aware Policy Optimization for Large Language Models

Why Attention Patterns Exist: A Unifying Temporal Perspective Analysis

SteadyThought: Mitigating LLM Under-Thinking via Thought-Level Preference Optimization

HiPO-MILP: Hierarchical Preference Optimization for MILP Solving

LogicTree: Improving Complex Reasoning of LLMs via Instantiated Multi-step Synthetic Logical Data

High-Performance Arithmetic Circuit Optimization via Differentiable Architecture Search

LaMPlace: Learning to Optimize Cross-Stage Metrics in Macro Placement

Differentiable Integer Linear Programming

Dynamic Configuration for Cutting Plane Separators via Reinforcement Learning on Incremental Graph

STNet: Spectral Transformation Network for Solving Operator Eigenvalue Problem

Mixture-of-Experts Operator Transformer for Large-Scale PDE Pre-Training

Accurate KV Cache Eviction via Anchor Direction Projection for Efficient LLM Inference

Computing Circuits Optimization via Model-Based Circuit Genetic Evolution

A Graph Enhanced Symbolic Discovery Framework For Efficient Logic Optimization

ROPO: Robust Preference Optimization for Large Language Models

AttentionPredictor: Temporal Patterns Matter for KV Cache Compression

OptiTree: Hierarchical Thoughts Generation with Tree Search for LLM Optimization Modeling

SymMaP: Improving Computational Efficiency in Linear Solvers through Symbolic Preconditioning

RoME: Domain-Robust Mixture-of-Experts for MILP Solution Prediction across Domains

Apollo-MILP: An Alternating Prediction-Correction Neural Solving Framework for Mixed-Integer Linear Programming

Accurate and Scalable Graph Neural Networks via Message Invariance

Learning Robust Representations with Long-Term Information for Generalization in Visual Reinforcement Learning

Accelerating Large Language Model Reasoning via Speculative Search

HyperTree Planning: Enhancing LLM Reasoning via Hierarchical Thinking

Accelerating Eigenvalue Dataset Generation via Chebyshev Subspace Filter

Knowledge Graph Finetuning Enhances Knowledge Manipulation in Large Language Models

ROPO: Robust Preference Optimization for Large Language Models

Boosting Multi-Domain Fine-Tuning of Large Language Models through Evolving Interactions between Samples

SymMaP: Improving Computational Efficiency in Linear Solvers through Symbolic Preconditioning

STNet: Spectral Transformation Network for Solving Operator Eigenvalue Problem

Accelerating Eigenvalue Dataset Generation via Chebyshev Subspace Filter

Common Neighbor Induced Message Passing for Inductive Link Prediction in Knowledge Graphs

Can Symbolic Regression of Boolean Functions Boost Logic Synthesis?

Benchmarking End-To-End Performance of AI-Based Chip Placement Algorithms

SymMaP: Improving Computational Efficiency in Linear Solvers through Symbolic Preconditioning

Searching Strengthens Large Language Models in Finding Bugs of Deep Learning Libraries

LLM4Solver: Large Language Model for Efficient Algorithm Design of Combinatorial Optimization Solver

SReNet: Spectral Refined Network for Solving Operator Eigenvalue Problem