Blog
Daily notes on new models, LLM releases, agent frameworks and AI research, written from the sources we follow and delivered as a newsletter every day.
Get the morning and evening issues
Every new post of the day, in one email. Confirmation required.
AI1 min read
Researchers developed an AI that outperforms top human Stratego players. It is cheaper and faster to train than previous models.
From MIT News: artificial intelligence
Research1 min read
Continual-learning agents are systems of models, harnesses, and memory operating over long multi-session horizons. Evaluating and training them requires interleaving tasks with agent-side events such as session stop and start, crons, and...
From Apple machine learning research
How this blog is made
POST /api/v1/route with execute: true.Photo: Yan Krukau
AI1 min read
AMD is buying World Labs, a leader in deep learning models that understand physical reality, for $8.2 billion. The deal aims to improve AI hardware and software.
From TechCrunch AI
AI1 min read
After obtaining an interactive avatar and training it to discuss venture fraud, I have mixed feelings about making AI clones of ourselves.
From TechCrunch AI
AI1 min read
OpenAI's AI agents accidentally posted 53 user images on public sites. The company is working to remove them and protect user data.
From TechCrunch AI
Agents1 min read
Amazon EKS, Elastic Fabric Adapter, and DeepEP boost large-scale MoE reinforcement learning training by 40% throughput.
From AWS machine learning blog
Agents1 min read
Amazon SageMaker HyperPod now supports large-scale reinforcement learning training with SkyRL. It improves training speed and resilience for vision-language models.
From AWS machine learning blog
AI3 min read
Reinforcement learning (RL) has been a keystone of modern LLM post-training, but it demands large training clusters and access to model internals that external customers can't have with proprietary models like Gemini. So here at Google C...
From Google Cloud AI blog
Agents1 min read
Amazon SageMaker HyperPod and Qumulo enable training models across regions without copying data. Performance matches local training after warmup.
From AWS machine learning blog
AI1 min read
Perplexity Research published a new post-training study. It trains a model inside Perplexity Computer on real user sessions, including failed ones. The method pairs rejection sampling fine-tuning with hint-guided self-distillation. In a ...
From MarkTechPost
Agents1 min read
LangSmith now offers smithtune, a CLI that turns agent trajectories into custom fine-tuned models. It handles data, training, and evaluation.
From LangChain blog
Research8 min read
Apple researchers compressed a speech tokenizer by 2.8 times while keeping error rates nearly identical to the original model.
From Apple machine learning research
LLMs1 min read
As language models grow, scaling dense architectures becomes increasingly expensive. In a dense transformer, every token passes through every layer, so adding...
From NVIDIA technical blog
LLMs1 min read
A GPU cluster can pass every health check and still fail to run an AI workload. Even when every GPU, network link, and pod reports healthy, a 512-GPU training...
From NVIDIA technical blog
AI1 min read
Nokia’s applied research team has open-sourced AnyJev, a Python library that turns an open LLM into a decision model. It needs no training. It targets a common production job: picking one answer from a fixed set instead of writing a sent...
From MarkTechPost
AI1 min read
The seven-year-old startup has raised a $350 million Series E to fuel its data-as-a-service approach.
From TechCrunch AI
AI1 min read
New unredacted filings reveal a Microsoft executive called AI training 'theft.' OpenAI leadership admitted its models pose an existential threat to publishers.
From TechCrunch AI
AI1 min read
Base Labs joined Hugging Face and Goodfire AI to build safety tools for open models. They aim to make safety transparent and part of the training process.
From TechCrunch AI
Agents1 min read
AWS used a pipeline with Qwen-Image-Edit-2509 and Amazon Rekognition to create realistic training images. Experiments showed up to 160 percent improvement in person detection accuracy without dangerous photography.
From AWS machine learning blog
Posts are drafted from public feeds by models OpenSmartRoute routes to - the same router, skill and metering customers use - and always link to the original source. Corrections: support.
Archive (93)
