An expert overview of post-training techniques for language models, covering the entire workflow from data generation and curation to advanced algorithms like Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Reinforcement Learning (RL), along with practical advice on evaluation and iteration.
This talk introduces Anaximander, a system designed to bridge the gap between traditional, GUI-driven Geographic Information System (GIS) workflows and modern, code-heavy machine learning practices. Anaximander integrates geospatial foundation models directly into QGIS, allowing experts to interactively orchestrate, run, and evaluate models for tasks like semantic segmentation and object detection on satellite imagery.
At Applied Compute, efficient Reinforcement Learning is critical for delivering business value. This talk explores the transition from inefficient synchronous RL to a high-throughput asynchronous 'Pipeline RL' system. The core challenge is managing 'staleness'—a side effect of in-flight weight updates that can destabilize training. The speakers detail their first-principles systems model, based on the Roofline model, used to simulate and find the optimal allocation of GPU resources between sampling and training, balancing throughput with algorithmic stability and achieving significant speedups.
Zevi Arnovitz, a non-technical Product Manager at Meta, shares his complete workflow for building and shipping sophisticated applications using AI tools like Cursor. He details a structured, multi-step process that leverages different AI models for specific tasks, including a novel "peer review" technique where models critique each other's code.
![Why Every Brain Metaphor in History Has Been Wrong [SPECIAL EDITION]](https://img.youtube.com/vi/pO0WZsN8Oiw/maxresdefault.jpg)
An exploration of scientific simplification, questioning the metaphors we use to understand the brain and intelligence. This summary delves into the tension between creating useful models and mistaking them for reality, featuring insights on the mind-as-software debate, the limits of prediction versus understanding, and the philosophical underpinnings of our quest for AGI.
Thomas Wolf, co-founder of Hugging Face, discusses his journey from physics to AI, the power of open-source models to accelerate innovation, the practical challenges of productionalizing AI demos, and why the biggest opportunities for founders now lie in the application layer on top of powerful foundation models.
Jeff Crume and Patrick Fussell from IBM's X-Force team share a real-world ethical hacking war story, demonstrating an attack from an 'assume breach' perspective. They break down how vulnerabilities in Identity and Access Management (IAM) and legacy systems can lead to a full compromise, starting from an insider threat and escalating to domain administrator privileges through advanced C2 attacks and lateral movement.
Nikesh Arora, CEO of Palo Alto Networks, shares his unconventional journey and leadership philosophy. He provides a masterclass in building a multi-platform company through strategic M&A, explains why founders should sometimes ignore customers, and reveals how to lead with conviction while managing imposter syndrome.
Stewart Butterfield, co-founder of Slack and Flickr, shares the product frameworks and leadership principles that guided his success. He delves into concepts like "utility curves" for feature investment, the "owner's delusion" in product design, and why focusing on "comprehension" is often more important than reducing friction. He also introduces powerful mental models for organizational effectiveness, such as combating "hyper-realistic work-like activities" and applying Parkinson's Law to team growth.
Sebastian Thrun, co-founder of Google's Moonshot Factory, recounts the early days of X and the Waymo self-driving car project. He shares insights into the unique management philosophy that fostered radical innovation, the ethical responsibilities of technologists, and his optimistic vision for the future of AI.
Douwe Kiela, CEO of Contextual AI, explains the evolution from basic RAG to "RAG 2.0", an end-to-end, trainable system. He argues that this system-level approach, which integrates optimized document parsing, retrieval, reranking, and grounded models, is superior to relying on massive context windows alone and is a fundamental tool for next-generation AI agents.
Push Kohli and Máté Balog from Google DeepMind discuss AlphaDev, an AI agent that uses large language models and evolutionary search to discover novel, more efficient algorithms for fundamental computer science problems, marking a significant step in AI's ability to generate creative and practical solutions.
Anish Agarwal and Raj Agrawal, co-founders of Traversal, discuss how their AI agents automate root cause analysis (RCA) for critical system failures. They detail their agent's architecture, which leverages causal inference and large-scale computation to systematically find the root cause in minutes, and argue that the rise of AI-generated code makes AI-powered debugging an essential capability for modern software engineering.
AI is being used to address critical information gaps in ecology. This summary covers how deep learning models like vision transformers and foundational models for sound are applied to map global forests, monitor deforestation, track species, and analyze bioacoustics to understand ecosystem health and animal communication.
Experts from Isomorphic Labs discuss how AI, particularly models like AlphaFold3, is revolutionizing drug discovery. They explain the shift from the slow, iterative process of traditional methods to a new era where generative AI can design novel molecules, optimize for multiple biological properties simultaneously, and pave the way for personalized medicine.
Get updates straight to your inbox. No spam, just useful content.