The Schedule

10:00-16:00 ET, July 13th

  • Opening Remarks [10:00]

  • (Session I) Foundations of Sparsity: From Mathematics to Neuroscience (Chair: Ari Morcos & Atlas Wang)

    • [10:10] Invited Talk 1: Dimitris Papailiopoulos (30 minute + 5min QA)

    • [10:45] Invited Talk 2: Surya Ganguli (30 minute + 5min QA)

    • [11:20] Spotlight 1: NISPA: Neuro-Inspired Stability-Plasticity Adaptation for Continual Learning in Sparse Networks (Mustafa Burak Gurbuz)

    • [11:30] Spotlight 2: Universality of Winning Tickets: A Renormalization Group Perspective (Will Redman)

    • [11:40] Spotlight 3: Lottery Tickets on a Data Diet: Finding Initializations with Sparse Trainable Networks (Brett W Larsen)

  • Poster Session [11:50-13:15]

  • Break [13:15-13:30]

  • (Session 2) Practice of Sparsity: Algorithms and Hardware (Chair: Jonathan Frankle & Decebal Mocanu)

    • [13:30] Invited Talk 1: Beidi Chen (30 minute + 5min QA)

    • [14:05] Invited Talk 2: Yanqi Zhou (30 minute + 5min QA)

    • [14:40] Spotlight 4: Multimodal Contrastive Learning with LIMoE: the Language-Image Mixture of Experts (Carlos Riquelme)

    • [14:50] Spotlight 5: FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness (Tri Dao)

    • [15:00] Spotlight 6: DRAGONN: Distributed Randomized Approximate Gradients of Neural Networks (Zhaozhuo Xu)

  • Panel [15:10-16:00] (Moderator: Jonathan Frankle)

    • w/ Dimitris Papailiopoulos, Beidi Chen, Yanqi Zhou and Surya Ganguli