Y Hacker News new | ask | show | jobs

user: pr337h4m created: 2022-08-20 karma: 4143 preetham<zero><four><zero><eight> @ the largest mail provider submissions: PerceptionBench: Evaluating Atomic Visual Perception in Multimodal LLMs 2 points | 0 comments Kimi K3 License 4 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments 0 points | 0 comments Clive Wearing 1 points | 0 comments 0 points | 0 comments 0 points | 0 comments Introducing ChatGPT (2022) 4 points | 0 comments Machine Studying 3 points | 0 comments Before Mythos, Satan: A 1990s Software Satanic Panic 4 points | 0 comments River AI 3 points | 0 comments Post-training speech models for better interactivity 2 points | 0 comments Invisible limitations on Claude Fable 5's effectiveness for frontier LLM dev 3 points | 0 comments Unitree: Voice‑driven, real‑time arbitrary action generation [video] 2 points | 0 comments 0 points | 0 comments Mythos set off a cybersecurity 'hysteria.' Experts say threat was already here 7 points | 2 comments Auto Exacto: Adaptive Quality Routing, on by Default 1 points | 0 comments Two Heads Are Better Than One: Async Knowledge Injection for Speech AI 2 points | 0 comments Exa and Google Partnership 4 points | 0 comments Amateur armed with ChatGPT solves an Erdős problem 796 points | 560 comments Software Is Speech: Why Regulators Cannot Invent the Missing Middlemen [pdf] 3 points | 0 comments You May Already Be Canadian 3 points | 0 comments The exam that broke society 2 points | 0 comments The Digital Imprimatur (2003) 4 points | 0 comments 0 points | 0 comments PostTrainBench: How well can AI agents post-train language models? 3 points | 0 comments Nemotron 3 Super: An Open Hybrid Mamba-Transformer Moe for Agentic Reasoning 2 points | 0 comments Gemma Needs Help 38 points | 1 comments