Hao Liang

Learning What Reinforcement Learning Can't: Interleaved Online Fine-Tuning for Hardest Questions

MC-LLaVA: Multi-Concept Personalized Vision-Language Model

Boosting RL-based Multimodal Reasoning via Difficulty Prior

Jarvis: Towards Personalized AI Assistant via Personal KV-Cache Retrieval

Small-Large Collaboration: Training-efficient Concept Personalization for Large VLM using a Meta Personalized Small VLM

UniCTokens-R1: Boosting Unified Personalization via Reinforcement Learning

GovBench: From Natural Language to Executable Pipelines, A New Benchmark for Data Governance Automation

UniCTokens: Boosting Personalized Understanding and Generation via Unified Concept Tokens

Facilitating Multi-turn Function Calling for LLMs via Compositional Instruction Tuning

QAEncoder: Towards Aligned Representation Learning in Question Answering System

BEATS: Optimizing LLM Mathematical Capabilities with BackVerify and Adaptive Disambiguate based Efficient Tree Search

MathScape: Evaluating MLLMs in Multi-modal Math Scenarios through a Hierarchical Benchmark

PAS: Plug-and-Play Prompt Augmentation System

Gradual Learning: Optimizing Fine-Tuning with Partially Mastered Knowledge in Large Language Models