Search papers, labs, and topics across Lattice.
Graduate School of Advanced Science and Engineering, Hiroshima University
7
0
8
4
Organizing demonstrations into a simple-to-complex structure can dramatically boost the efficiency and stability of robotic manipulation learning.
Integrating vision and radar data can drastically enhance the quality of point clouds, improving detection accuracy in challenging environments.
Time series language models can achieve up to 7.68脳 faster inference and improved performance by intelligently compressing tokens based on their information structure.
By leveraging frequency domain analysis, this approach significantly enhances the robustness of 3D perception systems against diverse driving conditions without needing target-domain samples.
Catastrophic forgetting in LLM fine-tuning with Evolution Strategies isn't irreversible or unique to ES, and can be largely avoided with a simple regularization technique.
LLMs actually *do* improve time series forecasting, especially for cross-domain generalization, overturning prior doubts with a massive 8-billion observation study.
Fine-tune quantized LLMs without backpropagation using QES, a method that achieves high-precision results at low-precision cost.
