Lei Zhang
8 papers · Latest:
Guide, Think, Act: Interactive Embodied Reasoning in Vision-Language-Action Models
GTA-VLA is an interactive Vision-Language-Action framework that uses user-provided spatial guidance to improve robot reasoning and robustness in embodied tasks.
Validity and Limits of Low Order Hybridization Expansion Approaches for Multi-Orbital Systems
Low-order hybridization expansion methods' accuracy in multi-orbital systems is limited by the least correlated orbital, which suppresses features.
Beyond Continuity: Simulation-free Reconstruction of Discrete Branching Dynamics from Single-cell Snapshots
Unbalanced Schrödinger Bridge (USB) reconstructs discrete branching cell dynamics from snapshots, integrating stochastic and birth-death events.
ALAS: Adaptive Long-Horizon Action Synthesis via Async-pathway Stream Disentanglement
ALAS uses dual-stream disentanglement for long-horizon human-scene interaction tasks, improving success and efficiency across domains.
CreatiParser: Generative Image Parsing of Raster Graphic Designs into Editable Layers
CreatiParser is a new generative model that parses raster graphic designs into editable text, background, and sticker layers for easy editing.
MasterSet: A Large-Scale Benchmark for Must-Cite Citation Recommendation in the AI/ML Literature
MasterSet is a new large-scale benchmark for identifying critical 'must-cite' papers in AI/ML literature, addressing a gap in existing citation recommendation systems.
Multimodal Reasoning with LLM for Encrypted Traffic Interpretation: A Benchmark
This paper introduces BGTD, a new benchmark, and mmTraffic, an LLM-based multimodal framework for explainable encrypted network traffic interpretation.
The Llama 3 Herd of Models
Llama 3 is a new family of large multilingual foundation models excelling in language, coding, reasoning, and multimodal tasks, rivaling GPT-4 in quality and offering extensive public releases.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.