Seminars

Reinforcement learning (RL) is a powerful general-purpose technique for automatically selecting actions that maximize rewards.

Joint work with Alexander Bentkamp, Simon Cruanes, Visa Nummelin, Stephan Schulz, Sophie Tourret, Petar Vukmirović, and Uwe Waldmann

Distributed training faces fundamental challenges from client heterogeneity in compute, memory, and network conditions. Existing approaches use staleness-dependent decay, per-client adjustments, or distance-weighted averaging, but often lack substantial convergence guarantees.

*Abstract*

In usual foundations like ZFC, we perform all of our mathematical constructions out of bare sets. The axiom of choice ensures these sets are "extremally disconnected" --- Cantor dust.

This Tech Talk will cover several key considerations and the practical demands of building enterprise grade software services.

*Abstract*: The denoising diffusion probabilistic model (DDPM) has become a cornerstone of generative AI.

Talk 1.

NetFridgeS: Enabling Dynamic Frequency Scaling on Network Switches through Carbon-Aware Routing

Talk 1.

NetFridgeS: Enabling Dynamic Frequency Scaling on Network Switches through Carbon-Aware Routing