Class Presentations by Students

Each student is expected to lead a lecture of his/her choice of topics, with the instructor's apporval. All students are required to meet the instructor on one-to-one basis to discuss the lecture materials in detail prior to the presentation. One week before the scheduled presentation, s/he will be expected to submit a draft version of the presentation materials and an initial treatment of the selected topics. The instructor will provide timely feedback about the pre-talk. Reading materials and/or discussion issues will be posted on the course web site, at least one day prior to each lecture. All class members will be expected to have read the listed readings, by the start of the relevant class.

Important Deadlines:

  • By October 9, 2026 - Choose a presentation topic and inform the instructor with a list of reading materials
  • One week before the presentation - Submit a draft of presentation materials
  • One lecture before the presentation - Hand out copies of reading materials, if not available online
  • One day before the presentation - Post the presentation materials on the web
  • Instructions for posting the lecture notes, reading materials, etc.


    Here is the list of topics to be presented by students in the chronological order:

  • Differentiable Meshing and Rendering (Richard Chen) (Tues, Oct 27, 2026)
  • Differentiable Meshing and Rendering (Yucheng Zu) (Tues, Oct 27, 2026)
  • Differentiable Real-to-Sim Reconstruction (Tues, Oct 29, 2026)
  • Reconstruction of Human-Object Interaction from Monocular Video (Tues, Oct 29, 2026)
  • Neural SDEs (Thur, Oct 29, 2026)
  • Gradients Through Discrete Decisions (Tues, Nov 10, 2026)
  • Differentiable Programming for Genomics: Learning Cell Dynamics with Neural SDEs (scDiffEq) (Tues, Nov 10, 2026)
  • Flow Matching: Foundations and Recent Advances (Tues, Nov 10, 2026)
  • Human Detection and Object Categorization based on WIFI CSI (Thur, Nov 12, 2026)
  • Visual Water Quality Monitoring and Auxiliary Modality Learning for Multimodal Data: Methods, Trends, and Opportunities (Thur, Nov 12, 2026)
  • Planning from Pixels: JEPA and LeWorldModel (Thur, Nov 12, 2026)
  • Understanding Catastrophic Forgetting during Sequential Fine-Tuning (Tues, Nov 17, 2026)
  • Continual learning for robot control under changing operation (Tues, Nov 17, 2026)
  • Learning Anticipatory Social Navigation from Human-Human Interactions Using Different Learning (Tues, Nov 17, 2026)
  • Learning Robot Navigation from Differentiable Planning Objectives (Thur, Nov 19, 2026)
  • What do Robots Need to Feel as They Walk (Thur, Nov 19, 2026)
  • Multisensory robot learning and the Affordance Problem (Thur, Nov 19, 2026)
  • Accelerating Robot Reinforcement Learning through Adaptive Rewards and Teaching Strategies (Tues, Nov 24, 2026)
  • Differentiable Model Predictive Control for Autonomous Vehicles with Nonlinear Tire Dynamics (Tues, Nov 24, 2026)
  • rone Swarms and Differentiable Programming (Tues, Nov 24, 2026)
  • Differentiable Trajectory Optimization (Tues, Dec 1, 2026)
  • Differentiable Learning of Adaptive Spoken Interaction (Tues, Dec 1, 2026)
  • Differentiable Ancient Language Reconstruction (Tues, Dec 1, 2026)
  • Can visual representation understand acoustic information? (Thur, Dec 3, 2026)
  • Automatic Prompt Optimization for LLM Agents (Thur, Dec 3, 2026)
  • Optimizing the Workflow of Multi-Agent Systems (Thur, Dec 3, 2026)
  • Optimizing Agentic Systems: From Textual Gradients to Reflective Evolution, focusing on TextGrad and GEPA (Tues, Dec 8, 2026)
  • Differentiation of Circuits (Tues, Dec 8, 2026)
  • Towards Differentiable Phenomenological Modeling: Applications in Automated Analysis of Semiconductor Quantum Dot Qubit Measurements (Tues, Dec 8, 2026)
  • Differentiation of Quantum Circuit Parameters for Noise Mitigation (Thur, Dec 10, 2026)
  • Exploring Role of Active-Gaze in Heterogeneous Agent Collaboration (Thur, Dec 10, 2026)
  • Learning to Accelerate Mathematical Algorithms Through Differentiable Programming (Thur, Dec 10, 2026)

  • Each student presentation will be graded based upon:

  • Advanced Preparation According to the Specification (40%)
  • Analysis and Discussion of the Materials (40%)
  • Style and Clarity of the Actual Presentation (20%)