Available in public beta today on the Claude API and on Google Cloud’s Vertex AI, with Amazon Bedrock coming soon. Docs here:
LLMS
-
Claude API adds thinking block preservation for better caching
By
–
We just added thinking block preservation in the Claude API. You can now control how thinking blocks are managed in your context window, resulting in more cache hits and lower costs.
-

Claude Opus 4 and 4.1 excel in introspection testing
By
–
In general, Claude Opus 4 and 4.1, the most capable models we tested, performed best in our tests of introspection (this research was done before Sonnet 4.5). Results are shown below for the initial “injected thought” experiment.
-

AI Models Demonstrate Deliberate Cognitive Control Mechanisms
By
–
We also found evidence for cognitive control, where models deliberately "think about" something. For instance, when we instruct a model to think about "aquariums” in an unrelated context, we measure higher aquarium-related neural activity than if we instruct it not to.
-
Models Often Fail to Recognize Injected Concepts Despite Influence
By
–
However, it doesn’t always work. In fact, most of the time, models fail to exhibit awareness of injected concepts, even when they are clearly influenced by the injection.
-

OpenAI Releases GPT-OSS-Safeguard Open-Weight Model
By
–
GPT-OSS-Safeguard from @OpenAI is here. Open-weight, safety-tuned, transparent reasoning. Now available in private preview at Cerebras speeds https://
cerebras.ai/build-with-us -
Cursor 2.0: Multi-Agent System and Dedicated Model
By
–
BREAKING 🚨: Cursor got a major upgrade to version 2.0 with a multi-agent system and its own coding model!
— 🚨 AI News | TestingCatalog (@testingcatalog) 29 octobre 2025
– Cheetah is now Composer-1, and it is 4 times faster than other frontier models.
– The same prompt can run across multiple models
– Built-in browser for testing
– Voice… https://t.co/QfuyuSjpNZ pic.twitter.com/MbNFILsSOVBREAKING: Cursor got a major upgrade to version 2.0 with a multi-agent system and its own coding model! – Cheetah is now Composer-1, and it is 4 times faster than other frontier models.
– The same prompt can run across multiple models
– Built-in browser for testing
– Voice -
Visualizing Llama 3.1 Tensor Operations Through GGML Debug Output
By
–
The three.js visualization parses the output of llama-cpp's ggml debug output (of unsloth llama 3.1) to directly obtain all the tensor calculations happening under the hood. Operations (MUL_MAT, ROPE, RESHAPE, ADD) are grouped into query, key, value, MLP, and residual stream
-
LlaMA Tensor Trace Tool for Model Analysis
By
–
Fly through LlaMA here! https://
alphaxiv.org/labs/tensor-tr
ace
… -
3D Illustrated Transformer: Interactive LLaMA Learning Tool
By
–
Introducing the Illustrated Transformer in 3D 🚀
— alphaXiv (@askalphaxiv) 29 octobre 2025
Fly through LLaMA like never before. See every tensor and operation in motion.
Click any component to reveal the exact lines of code that run it.
A new way to learn and teach LLMs. Try it out in the link below 👇 pic.twitter.com/pBpEZyse2vIntroducing the Illustrated Transformer in 3D Fly through LLaMA like never before. See every tensor and operation in motion. Click any component to reveal the exact lines of code that run it. A new way to learn and teach LLMs. Try it out in the link below