weipeng chen

Exploring the Design Space of Visual Context Representation in Video MLLMs

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

Needle In A Video Haystack: A Scalable Synthetic Evaluator for Video MLLMs

Extracting and Transferring Abilities For Building Multi-lingual Ability-enhanced Large Language Models

SysBench: Can LLMs Follow System Message?

Beyond Filtering: Adaptive Image-Text Quality Enhancement for MLLM Pretraining

Optimizing Inference-Time Reasoning in LLMs via Retrieval-Augmented Reflection

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

FB-Bench: A Fine-Grained Multi-Task Benchmark for Evaluating LLMs' Responsiveness to Human Feedback

ShortGPT: Layers in Large Language Models are More Redundant Than You Expect

PAS: Plug-and-Play Prompt Augmentation System

Base of RoPE Bounds Context Length

Exploring Context Window of Large Language Models via Decomposed Positional Vectors