Fengli Xu

FingerTip 20K: A Benchmark for Proactive and Personalized Mobile LLM Agents

Reinforcement Learning Fine-Tuning Enhances Activation Intensity and Diversity in the Internal Circuitry of LLMs

TravelReasoner: Reasoning-Augmented Travel Survey Simulations with Large Reasoning Models

SciNetBench: A Relation-Aware Benchmark for Scientific Literature Retrieval Agents

Route-and-Reason: Scaling Large Language Model Reasoning with Reinforced Model Router

Towards Generalizable LLM Multi-Agent System: Identifying Collective Intelligence Factor in LLM Agent Groups

Deep Ideation: Designing LLM Agents to Generate Novel Research Ideas on Scientific Concept Network

Controllable Test-Time Scaling via Sparse Autoencoder‑Based Reasoning Steering

Eliciting Human-like Social Reasoning in Large Language Models

AgentSquare: Automatic LLM Agent Search in Modular Design Space

Token Signature: Predicting Chain-of-Thought Gains with Token Decoding Feature in Large Language Models

OpenCity: A Scalable Platform to Simulate Urban Activities with Massive LLM Agents