OpenAI's Democratic Inputs to AI💡 // MiniCPM-2B Language Model 🤯 // V-IRL Platform for Real-World AI 🌍

OpenAI's plan to involve the public in AI governance through "Democratic Inputs to AI" is a promising idea. MiniCPM-2B, an edge-side large language model, outperforms many other models on comprehensive benchmarks. DeepSeekMath 7B achieved a score of 51.7% on the MATH benchmark for mathematical reasoning. V-IRL is a platform that enables AI agents to interact with the real world, opening up possibilities for practical applications such as disaster response and environmental monitoring. Contact:  [email protected] Timestamps: 00:34 Introduction 01:28 Inside OpenAI’s Plan to Make AI More ‘Democratic’ 03:46 MiniCPM: Unveiling the Potential of End-side Large Language Models 05:06 Beyond Token Prediction: the post-Pretraining journey of modern LLMs 06:56 Fake sponsor 08:36 DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models 10:19 Open RL Benchmark: Comprehensive Tracked Experiments for Reinforcement Learning 12:13 V-IRL: Grounding Virtual Intelligence in Real Life 13:54 Outro