Feng Wu

Latest

• Accurate KV Cache Eviction via Anchor Direction Projection for Efficient LLM Inference
• Benchmarking End-To-End Performance of AI-Based Chip Placement Algorithms
• Dynamic Configuration for Cutting Plane Separators via Reinforcement Learning on Incremental Graph
• High-Performance Graph System Optimization via Differentiable Architecture Search
• Accelerating Large Language Model Reasoning via Speculative Search
• HyperTree Planning: Enhancing LLM Reasoning via Hierarchical Thinking
• ROPO: Robust Preference Optimization for Large Language Models
• Accelerate Presolve in Large-Scale Linear Programming via Reinforcement Learning
• Knowledge Graph Finetuning Enhances Knowledge Manipulation in Large Language Models
• A Graph Enhanced Symbolic Discovery Framework For Efficient Logic Optimization
• Accurate and Scalable Graph Neural Networks via Message Invariance
• Apollo-MILP: An Alternating Prediction-Correction Neural Solving Framework for Mixed-Integer Linear Programming
• Computing Circuits Optimization via Model-Based Circuit Genetic Evolution
• Differentiable Integer Linear Programming
• LaMPlace: Learning to Optimize Cross-Stage Metrics in Macro Placement
• Long-term Feature Extraction via Frequency Prediction for Efficient Reinforcement Learning
• MILP-StuDio: MILP Instance Generation via Block Structure Decomposition
• Towards Next-Generation Logic Synthesis: A Scalable Neural Circuit Generation Framework
• Label Deconvolution for Node Representation Learning on Large-scale Attributed Graphs against Learning Bias
• A Circuit Domain Generalization Framework for Efficient Logic Synthesis in Chip Design
• A Hierarchical Adaptive Multi-Task Reinforcement Learning Framework for Multiplier Circuit Design
• Coarse-to-Fine Highlighting: Reducing Knowledge Hallucination in Large Language Models
• Reinforcement Learning within Tree Search for Fast Macro Placement
• Towards General Algorithm Discovery for Combinatorial Optimization: Learning Symbolic Branching Policy from Bipartite Graph
• Learning to Cut via Hierarchical Sequence/Set Model for Efficient Mixed-Integer Programming
• Accelerating Data Generation for Neural Operators via Krylov Subspace Recycling (Spotlight)
• Rethinking Branching on Exact Combinatorial OPtimization Solver: The First Deep Symbolic Discovery Framework
• A Deep Instance Generative Framework for MILP Solvers Under Limited Data Availability (Spotlight)
• State Sequences Prediction via Fourier Transform for Representation Learning (Spotlight)
• De Novo Molecular Generation via Connection-aware Motif Mining
• Learning Cut Selection for Mixed-Integer Linear Programming via Hierarchical Sequence Model
• Modeling Diverse Chemical Reactions for Single-step Retrosynthesis via Discrete Latent Variables
• Compressing Deep Graph Neural Networks via Adversarial Knowledge Distillation
• Learning Task-relevant Representations for Generalization via Characteristic Functions of Reward Sequence Distributions
• Duality-Induced Regularizer for Semantic Matching Knowledge Graph Embeddings
• Rethinking Graph Convolutional Networks in Knowledge Graph Completion
• ConE: Cone Embeddings for Multi-Hop Reasoning over Knowledge Graphs
• Deep Cognitive Reasoning Network for Multi-hop Question Answering over Knowledge Graphs
• Topology-Aware Correlations Between Relations for Inductive Link Prediction in Knowledge Graphs
• Self-Adaptive Embedding for Few-shot Classification by Hierarchical Attention