Jinqiao Wang

Scaling Linear Attention Capacity with Sparse State Expansion

Connecting Where You Look With What You Understand: Trajectory-Driven Localized Understanding for Interactive Vision-Language Models

Characterizing and Mitigating Reasoning Drift in Large Language Models

Improving Complex SQL Generation for Text-to-SQL by Addressing Semantic Blind Spots in Pending SQL Components

Dichotomous Diffusion Policy Optimization

ReST-KV: Robust KV Cache Eviction with Layer-wise Output Reconstruction and Spatial-Temporal Smoothing

GThinker: Towards General Multimodal Reasoning via Cue-Guided Rethinking

VFaith: Do Large Multimodal Models Really Reason on Seen Images Rather than Previous Memories?

Understand, Think, and Answer: Advancing Visual Reasoning with Large Multimodal Models

Vision-R1: Evolving Human-Free Alignment in Large Vision-Language Models via Vision-Guided Reinforcement Learning

PhysVLM-AVR: Active Visual Reasoning for Multimodal Large Language Models in Physical Environments

FOCUS: Unified Vision-Language Modeling for Interactive Editing Driven by Referential Segmentation

Systematic Outliers in Large Language Models

Recurrent Context Compression: Efficiently Expanding the Context Window of LLM