Focal Interest
Focal Interest

Bottleneck

  1. Low-power gating system reduces data throughput in implantable BCI
    NeuroGator gates brain signal processing in implantable devices, reducing power consumption and data throughput by 82% while maintaining classification accuracy for next-generation BCIs.
  2. CORE: Data Augmentation for Link Prediction via Information Bottleneck
    CORE applies Information Bottleneck principles to augment graph data for link prediction, simultaneously recovering missing edges and reducing noise to enhance model robustness.
  3. P2P: From Prompt to Prototype - Functional UI Design with LLMs and MCP
    Tutorial addressing the implementation gap in HCI research through LLM and MCP-enabled functional UI design, bridging conceptual design and deployable interfaces while maintaining accessibility.