🎥 Random Samples
Random Samples is a weekly seminar series that airs live on our Red Hat YouTube page.
📹 Instance-Adaptive Inference-Time Scaling with Calibrated Process Reward Models
👤 Speaker: Young Jin Park
📹 Hopscotch: Discovering and Skipping Redundancies in Language Models
👤 Speaker: Mustafa Eyceoz
📹 Grounding Feedback is All You Need: Aligning Small Vision-Language Models
👤 Speaker: Giorgio Giannone
📹 On scalable RL in the era of agentic LLMs
👤 Speaker: Aldo Pareja
📹 LLM Meets Cache: From Application to Architecture
👤 Speaker: Shuhang Lin
📹 Accelerating LLM Knowledge Learning and Unlearning Research via Unified Frameworks
👤 Speaker: Wenlong Zhao
📹 Towards Combinatorial Interpretability of Neural Computation
👤 Speaker: Nir Shavit
📹 Synthetic Data Generation via SDG-Hub
👤 Speaker: Abhishek Bhandwaldar, Shivchander Sudalairaj, Akash Srivastava
📹 Continual Post-Training
👤 Speaker: Nikhil Nayak, Krishnateja Killamsetty
📹 State of LLM Compression from Research to Production
👤 Speaker: Mark Kurtz
📹 Activation-Informed Merging of Large Language Models
👤 Speaker: Amin Heyrani Nobari & Kaveh Alim
📹 Scaling Inference Time Scaling: Subspace-orthogonal KV Cache Quantization
👤 Speaker: Hao Wang, Ligong Han
📹 Practitioner's Guide to Instruction Tuning
👤 Speaker: Aldo Pareja
🎙️ No Math AI
No Math AI is a monthly podcast designed for the AI community and enthusiasts who want to better understand how AI research translates into real-world business impact—without the complex equations.
Hosted by Dr. Akash Srivastava (Red Hat) and Isha Puri (MIT), this series distills critical AI concepts into practical takeaways to help practitioners, businesses, and curious minds accelerate adoption with confidence.
📹 Inference Time Scaling for Enterprises
👤 Speaker: Matt Hicks, Chris Wright
🎧 Listen on Spotify
📹 Generative Optimization in Engineering Design
👤 Speaker: Dr. Faez Ahmed
🎧 Listen on Spotify
📹 Inference-time scaling: How small models beat the big ones
👤 Speaker: Shivchander Sudalairaj, GX Xu, Kai Xu
🎧 Listen on Spotify
🎬 Other Select Talks
📹 Language Model Post-Training in 2025: an Overview of Customization Options Today
👤 Speaker: Mustafa Eyceoz
📹 SDG_Hub: An Open-Source Toolkit for Synthetic Data Generation & LLM Customization
👤 Speaker: Shivchander Sudalairaj & Abhishek Bhandwaldar
📹 Unlocking Smarter AI with Inference-Time Scaling
👤 Speaker: Kai Xu & GX Xu
📹 Sculpting Subspaces: Constrained Full Fine-Tuning for Continual Learning in LLMs
👤 Speaker: Nikhil Nayak