跳到主要内容

Research

"The future is here — just unevenly distributed." (William Gibson)

Research is our foundation. We transform cutting-edge research into accessible tools, applicable methodologies or usable systems.

Enhancing Cross-task Transfer of Large Language Models via Activation SteeringModel2025-07-17QuestA: Expanding Reasoning Capacity in LLMs via Question AugmentationSystem2025-07-17Toward Real-World Table Agents: Capabilities, Workflows, and Design Principles for LLM-based Table IntelligenceSystem2025-07-14ABench-Physics: Benchmarking Physical Reasoning in LLMs via High-Difficulty and Dynamic Physics ProblemsSystem2025-07-07Why Do Open-Source LLMs Struggle with Data Analysis? A Systematic Empirical StudyResearch2025-06-24Ring-lite: Scalable Reasoning via C3PO-Stabilized Reinforcement Learning for LLMsModel2025-06-17AutoMind: Adaptive Knowledgeable Agent for Automated Data ScienceResearch2025-06-12Ming-Omni: A Unified Multimodal Model for Perception and GenerationModel2025-06-11Right Is Not Enough: The Pitfalls of Outcome Supervision in Training LLMs for Math ReasoningResearch2025-06-07AREAL: A Large-Scale Asynchronous Reinforcement Learning System for Language ReasoningSystem2025-06-04