Zheng Liu

MR$^2$-Bench: Going Beyond Matching to Reasoning in Multimodal Retrieval

Open Data Synthesis for Deep Research

EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling

VideoExplorer: Boosting Long Video Understanding with Dynamic Temporal Grounding

Retro*: Optimizing LLMs for Reasoning-Intensive Document Retrieval

Reinforcing Agentic Search Via Reward Density Optimization

Taming Data Chaos: Agentic Knowledge Warehousing for Contextual Intelligence

TimeScope: Towards Task-Oriented Temporal Grounding In Long Videos

EfficientLLM: Unified Pruning-Aware Pretraining for Auto-Designed Edge Language Models

Scent of Knowledge: Optimizing Search-Enhanced Reasoning with Information Foraging

Long Context Compression with Activation Beacon

SpikeLLM: Scaling up Spiking Neural Network to Large Language Models via Saliency-based Spiking

Making Text Embedders Few-Shot Learners

MMTEB: Massive Multilingual Text Embedding Benchmark

Feint and Attack: Attention-Based Strategies for Jailbreaking and Protecting LLMs

Elephant in the Room: Unveiling the Pitfalls of Human Proxies in Alignment