Large Language Models
Jun 28, 2026

Resumo do dia
Airwallex hits $11 billion valuation with $320 million raise as fintech pushes into finance run by AI agents. ChatGPTはグーグル検索の10倍電力消費…丸紅の米天然ガス買収が映すAIインフラの「電力不足」. I shrank a transformer until every number fitted on the screen and made the weights editable [R]
Principais notícias
- 1
Airwallex hits $11 billion valuation with $320 million raise as fintech pushes into finance run by AI agents
Airwallex hits $11 billion valuation with $320 million raise as fintech pushes into finance run by AI agents
- 2
ChatGPTはグーグル検索の10倍電力消費…丸紅の米天然ガス買収が映すAIインフラの「電力不足」
ChatGPTはグーグル検索の10倍電力消費…丸紅の米天然ガス買収が映すAIインフラの「電力不足」
- 3
I shrank a transformer until every number fitted on the screen and made the weights editable [R]
I shrank a transformer until every number fitted on the screen and made the weights editable [R]
- 4
MathFormer: Testing whether symbolic math is pattern matching or reasoning [D]
MathFormer: Testing whether symbolic math is pattern matching or reasoning [D]
- 5
Evaluating long-term memory limits in stateless LLM chatbots — feedback needed [D]
Evaluating long-term memory limits in stateless LLM chatbots — feedback needed [D]
- 6
Hiding messages in the least significant mantissa bits of fine-tuned ONNX model weights [P]
Hiding messages in the least significant mantissa bits of fine-tuned ONNX model weights [P]
Fontes
- Airwallex hits $11 billion valuation with $320 million raise as fintech pushes into finance run by AI agents
- ChatGPTはグーグル検索の10倍電力消費…丸紅の米天然ガス買収が映すAIインフラの「電力不足」
- I shrank a transformer until every number fitted on the screen and made the weights editable [R]
- MathFormer: Testing whether symbolic math is pattern matching or reasoning [D]
- Evaluating long-term memory limits in stateless LLM chatbots — feedback needed [D]
- Hiding messages in the least significant mantissa bits of fine-tuned ONNX model weights [P]
- NagaTranslate: Building a translation and voice pipeline for low-resource Nagaland creoles (Whisper, VITS, LLMs) [P]
- Built an LLM training framework that actually runs on older GPUs without crashing [P]
- Prompt injection is exploiting enterprise AI's biggest design flaws by targeting agents, RAG pipelines and model routers
- Do LLMs Have Desires?
Share this with a friend
Send today's roundup to anyone who wants to keep up.
Get daily AI news free with AIToday
200+ AI sources, summarized in 1 minute. Email / LINE / Slack.
Sign up free