Lanling Tang
Papers
3
Total Citations
43
H-Index
2
About
Lanling Tang is at the forefront of integrating large language models (LLMs) with robotics, specializing in complex, long-horizon task planning. Their major contributions center on overcoming the limitations of LLMs in generating executable, multi-step robotic plans. Tang introduced **FLTRNN (Faithful Long-Horizon Task Planning)**, a method that enhances plan reliability by effectively managing the extensive context—instructions and demonstrations—needed for intricate tasks, achieving 26 citations since 2024. Building on this, Tang developed **MLDT (Multi-Level Decomposition)**, a framework that breaks down complex robotic tasks into manageable sub-problems, enabling open-source LLMs to perform robust planning without relying on proprietary models. With a combined 17 citations for their MLDT work, Tang’s research is pivotal in making advanced robotic planning more accessible and faithful. Their work addresses a critical bottleneck: ensuring that LLM-generated plans are not only coherent but also practically executable over long horizons. Tang’s achievements are shaping the next generation of autonomous robotics, where AI-driven reasoning seamlessly translates into real-world action.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3