Jure Leskovec

Multiplayer Nash Preference Optimization

Relational Transformer: Toward Zero-Shot Foundation Models for Relational Data

Optimas: Optimizing Compound AI Systems with Globally Aligned Local Rewards

TGM: A Modular and Efficient Library for Machine Learning on Temporal Graphs

Relational Graph Transformer

RelDiff: Relational Data Generative Modeling with Graph-Based Diffusion Models

RFG: Test-Time Scaling for Diffusion Large Language Model Reasoning with Reward-Free Guidance

Position: The Hidden Costs and Measurement Gaps of Reinforcement Learning with Verifiable Rewards

Uncalibrated Reasoning: GRPO Induces Overconfidence for Stochastic Outcomes

CollabLLM: From Passive Responders to Active Collaborators

Holographic Node Representations: Pre-training Task-Agnostic Node Embeddings

Energy-Based Diffusion Language Models for Text Generation

BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments

Zero-Shot Generalization of GNNs over Distinct Attribute Domains

TabDiff: a Mixed-type Diffusion Model for Tabular Data Generation

ContextGNN: Beyond Two-Tower Recommendation Systems

Zero-Shot Generalization of GNNs over Distinct Attribute Domains

Automated Hypothesis Validation with Agentic Sequential Falsifications

Learning Efficient Positional Encodings with Graph Neural Networks

Surface-based Peptide Design with Multi-modal Flow Matching

RelGNN: Composite Message Passing for Relational Deep Learning