Browse Lessons

Sponsored Session: Lightning Talk: Accelerated Software for a Post-Moore World - Jay Dawani

No GPU Left Behind: Scaling Online LLM Training With... - Mert Toslali & Yu Chin Fabian Lim

Lightning Talk: FlexCP: FlexAttention for Long Sequence Training... - Xilun Wu & Chien-Chin Huang

Breaking Heterogeneity Barriers: Unified Cloud-to-Robot AI System SW Stack for... - Yonghua Lin

The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic... - Yutaro Yamada

Lightning Talk: Flex Attention for Inference - Boyuan Feng & Driss Guessous, Meta

Agent2Agent Protocol and the Future of Agent Collaboration - Chitra Venkatesh, Google

Building and Sharing Reliable AI Agents with BeeAI - Sandi Besen, IBM

Lightning Talk: Hardware-Aware Python Packages ~ PyTorch and... - Jonathan Dekhtiar & Eli Uriegas

Sponsored Session: Amazingly Fast and Incredibly Scalable Inference... - Harry Kim & Laikh Tewari

Keynote: Welcome & Opening Remarks - Matt White, Executive Director, PyTorch Foundation

Coding with AI - Chip Huyen, Tep Studio

An Open Source Post-Training Stack: Kubernetes + Ray + PyTorch + vLLM - Robert Nishihara, Anyscale

Mapping the Open Model Landscape in 2025 - Nathan Lambert, Ai2

Thunder: Distribute and Optimize Your PyTorch Models With Zero... - Luca Antiga & Thomas Viehmann

Race to Superintelligence: Building Low-Friction AI Infrastructure for a Heterogen... - Ajit Mathews

Holistic Evaluation of Language Models (HELM) - Yifan Mai, Stanford University

Keynote: Framing the Frontier of Machine Intelligence - Joe Spisak, Meta

The Four Pillars of Large-Scale Decentralized Pre-Training: Methods, Bottlenecks... - Bowen Peng

Keynote: Scaling Intelligence: Infrastructure as the Foundation - Joe Spisak, Product Director, Meta

Panel Discussion: Are We Measuring Intelligence or Just Benchmarks?

Measuring Intelligence in the Wild: Advances in Preference-Based Evaluation -Anastasios Angelopolous

Mojo + PyTorch: A Simpler, Faster Path To Custom Kernels - Spenser Bauman, Modular

Scaling an Open Environments Ecosystem for Reinforcement Learning - Will Brown, Prime Intellect

Accelerating Open-Source RL and Agentic Inference with vLLM - Michael Goin, Red Hat | vLLM

Unlocking Performance: Harnessing LLMs To Streamline GPU Kernel Development in... - Jiannan Wang

Environments and Benchmarks for Cybersecurity Agents - Andy Zhang, Stanford University

Panel Discussion: The AI Kernel... - Robert Lange, Simran Arora, Nathan Lambert, Mark Saroufim

AGNTCY: The Open Source Project for Building Multi-Agent Systems - John Parello, Cisco

Scaling Agentic Intelligence from Pre-Training to RL - Aakanksha Chowdery

The Model Context Protocol: Connecting AI to Everything - Adam Jones, Anthropic

The Future Is Tiled: Using CuTile & TileIR To Write Portable, High-performance GPU...- Jared Roesch

mxfp8, mxfp4, nvfp4 formats and applications in PyTorch - Vasily Kuznetsov & Driss Guessous, Meta

Panel Discussion: Agents in the Wild: Real-World Deployments and Lessons Learned

Keynote: Noam Brown, Research Scientist at OpenAI in Conversation with Joe Spisak, Meta

Can Reinforcement Learning Lead to AGI? - Daniel Han, Unsloth

End-to-End Optimizing Multi-Turn RL and High-Performance Inference in Agents with... - Chenyang Zhao

Keynote: Closing Remarks - Matt White, Executive Director, PyTorch Foundation

Panel Discussion: Training and Inference at Planet Scale

Weaver: Shrinking the Generation-Verification Gap with Weak Verifiers - Jon Saad-Falcon

Panel Discussion: Beyond the Leaderboard... - Jeremy Howard, Rebecca Qian, Haifeng Xu, Shishir Patil

Sponsored Session: Everything Everywhere all at Once: vLLM...- Brittany Rockwell & Shireen Kheradpey

Keynote: Building ML Systems Foundations at the Age of AI - Tianqi Chen, NVIDIA & CMU

SkyRL: A Modular RL Library for LLM Agents - Sumanth Hegde, Anyscale & Tyler Griggs, UC Berkeley

ExecuTorch 1.0: General Availability Status for Mobile and Embedded...- Mergen Nachin & Cemal Bilgin

From Model To Trust: Building Upon Tamper-proof ML Metadata Records - Mihai Maruseac, Google

Keynote: Robot Foundation Models - Sergey Levine

Keynote: Olmo-Thinking: Training a Fully Open Reasoning Model - Nathan Lambert