Siddhant Gupta - Introduction to Transformers and evolution of large language models

Andrei Panferov - Quartet: Native FP4 training can be optimal for large language models

Tony Silveti Falls - Training neural networks at any scale

Nils Reimers - Introduction to Embeddings & Retrieval

Katrina Lawrence - ML Math Refresher

Michael Noukhovitch - Asynchronous RLHF Faster and More Efficient Off Policy RL for Language Models

Ambroise Odonnat - Large Language Models as Markov Chains

Andrej Jovanovic -Communication efficient training for foundation models through federated learning

Cohere Labs - Regional Africa welcomes Asio Evelyn Patra

Kumar Ashutosh - LLMs can see and hear without any training

Arion Das - SweEval: Do LLMs Really Swear?

Stella Li Spurious Rewards Rethinking Training Signals in RLVR

Piotr Nawrot - The Sparse Frontier Sparse Attention Trade offs in Transformer LLMs

Vallijah Subasri - The Potential for Near Term AI Risks to Evolve into Existential Threats in Health

Karolis Jucys - Interpretability in Action Exploratory Analysis of VPT, a Minecraft Agent

Jianwei Yang - Magma: A Foundation Model for Multimodal AI Agents

Frederik Kunstner - Heavy tailed imbalance and why Adam outperforms gradient descent on language mod

Betty Kyalo - Shaping Africa's Ethical AI Governance Future

Eugene Vinitsky - Robust Autonomy Emerges from Self Play

Seulki Park - Visually Consistent Hierarchical Image Classification

2025 Expedition Aya Closing Ceremony

Dr Hicham Badri - Optimizing Linear Layers for Faster Inference

Karan Dalal - One Minute Video Generation with Test Time Training

Leonard Bauersfeld - Champion level Drone Racing using Deep Reinforcement Learning

Jeremy Bernstein - Depths of First Order Optimization

Eli Lifland - Forecasting the Future - Eli Lifland on AI 2027 and What Comes Next

Aya Vision Lightning Talks and AMA

Konstantin Schuerholt - Weight Space Learning Treating Neural Network Weights as Data

Esben Kran - Security Startups to Preserve Human Agency in an Age of AGI

Azeez Saheed Ayanniyi - YarnGPT the building process