Distributional reinforcement learning in prefrontal cortex
Google DeepMind posted “Distributional reinforcement learning in prefrontal cortex , Google DeepMind.” A quick update for the record.
Artificial Intelligence Blogs
Short posts on models, agents, APIs, and the companies building them
Search results: “Agents & tools”
116 entriesDecember 2025
Scalable Neural Network Kernels
Google DeepMind posted “Scalable Neural Network Kernels , Google DeepMind.” A quick update for the record.
TIPS: Text-Image Pretraining with Spatial awareness
Google DeepMind posted “TIPS: Text-Image Pretraining with Spatial awareness , Google DeepMind.” A quick update for the record.
Learning Planning-compatible Cognitive Maps with Transformers in PartiallyObserved Environments
Google DeepMind posted “Learning Planning-compatible Cognitive Maps with Transformers in PartiallyObserved Environments , Google DeepMind.” A quick update for the record.
Michelangelo: Long Context Evaluations Beyond Haystacks via Latent Structure Queries
Google DeepMind posted “Michelangelo: Long Context Evaluations Beyond Haystacks via Latent Structure Queries , Google DeepMind.” A quick update for the record.
Neural Fields as Distributions: Signal Processing Beyond Euclidean Space
Google DeepMind posted “Neural Fields as Distributions: Signal Processing Beyond Euclidean Space , Google DeepMind.” A quick update for the record.
Holistic Safety and Responsibility Evaluations of Advanced AI Models
Google DeepMind posted “Holistic Safety and Responsibility Evaluations of Advanced AI Models , Google DeepMind.” A quick update for the record.
Unsupervised Keypoints with Stable Diffusion
Google DeepMind posted “Unsupervised Keypoints with Stable Diffusion , Google DeepMind.” A quick update for the record.
Feature Likelihood Divergence: Evaluating the Generalization of Generative Models Using Samples
Google DeepMind posted “Feature Likelihood Divergence: Evaluating the Generalization of Generative Models Using Samples , Google DeepMind.” A quick update for the record.
Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA
Google DeepMind posted “Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA , Google DeepMind.” A quick update for the record.
Advancing Biomedical Understanding with Multimodal Gemini
Google DeepMind posted “Advancing Biomedical Understanding with Multimodal Gemini , Google DeepMind.” A quick update for the record.
A Human-Inspired Reading Agent with Gist Memory of Very Long Contexts
Google DeepMind announced “A Human-Inspired Reading Agent with Gist Memory of Very Long Contexts , Google DeepMind.” This one looks larger than a routine product note.
OpenAI co-founds Agentic AI Foundation, donates AGENTS.md
OpenAI posted “OpenAI co-founds Agentic AI Foundation, donates AGENTS.md.” A quick update for the record.
Anthropic acquires Bun as Claude Code reaches $1B milestone
Anthropic posted “Anthropic acquires Bun as Claude Code reaches $1B milestone.” A quick update for the record.
Inside Mirakl's agentic commerce vision
OpenAI posted “Inside Mirakl's agentic commerce vision.” A quick update for the record.
November 2025
Effective harnesses for long-running agents
Anthropic posted “Effective harnesses for long-running agents.” A quick update for the record.
OpenAI releases shopping research in ChatGPT
OpenAI announced “Introducing shopping research in ChatGPT.” This one looks larger than a routine product note.
Predictability and Surprise in Large Generative Models
Anthropic posted “Predictability and Surprise in Large Generative Models.” A quick update for the record.
Collective Constitutional AI: Aligning a Language Model with Public Input
Anthropic posted “Collective Constitutional AI: Aligning a Language Model with Public Input.” A quick update for the record.
Claude’s Character
Anthropic posted “Claude’s Character.” A quick update for the record.
Toy Models of Superposition
Anthropic posted “Toy Models of Superposition.” A quick update for the record.
Alignment Research
Anthropic posted “Alignment Research.” A quick update for the record.
Building more with GPT-5.1-Codex-Max
OpenAI announced “Building more with GPT-5.1-Codex-Max.” This one looks larger than a routine product note.
GPT-5.1-Codex-Max System Card
OpenAI announced “GPT-5.1-Codex-Max System Card.” This one looks larger than a routine product note.
Making Claude Code more secure and autonomous with sandboxing
Anthropic posted “Making Claude Code more secure and autonomous with sandboxing.” A quick update for the record.
October 2025
we built OWL, the new architecture behind our ChatGPT-based browser, Atlas
OpenAI posted “How we built OWL, the new architecture behind our ChatGPT-based browser, Atlas.” A quick update for the record.
OpenAI releases Aardvark: OpenAI’s agentic security researcher
OpenAI announced “Introducing Aardvark: OpenAI’s agentic security researcher.” This one looks larger than a routine product note.
September 2025
Claude 3.7 Sonnet and Claude Code
Anthropic announced “Claude 3.7 Sonnet and Claude Code.” This one looks larger than a routine product note.
Claude is now generally available in Xcode
Anthropic posted “Claude is now generally available in Xcode.” A quick update for the record.
August 2025
Claude Code and new admin controls for business plans
Anthropic posted “Claude Code and new admin controls for business plans.” A quick update for the record.
July 2025
OpenAI launches ChatGPT agent
OpenAI launched ChatGPT agent on July 17, 2025, allowing the system to operate tools and complete multi-step tasks.
June 2025
OpenAI releases the Agents SDK for TypeScript
The agent orchestration SDK expanded beyond Python to TypeScript.
May 2025
Anthropic API adds Files, MCP, and longer prompt caching
A Files API, managed MCP connector, and one-hour prompt cache arrived together.
Anthropic API adds code execution
Claude gained a sandboxed Python execution tool for analysis and visualizations.
Responses API adds background mode
Long-running responses could execute asynchronously with improved reliability.
Responses API adds Code Interpreter and image tools
OpenAI expanded Responses with code execution, image generation, and improved file search.
Responses API adds remote MCP servers
OpenAI let Responses connect directly to remote Model Context Protocol servers.
Google announces MCP support for the Gemini API
Gemini SDK tooling began supporting the Model Context Protocol for external tools and data.
Gemini API adds URL context
Developers could pass links for Gemini to use as context in a request.
Gemini API adds asynchronous function calling
Long-running tools could return results later without blocking the active model session.
Anthropic adds web search to the API
Claude API calls could retrieve current web information with citations.
March 2025
Anthropic cuts token use for tool calls
Cache-aware rate limits and token-efficient tool use reduced cost and output tokens.
OpenAI releases the Agents SDK
The new SDK added orchestration, handoffs, guardrails, and tracing for agent workflows.
OpenAI launches the Responses API
A new API primitive combined model responses with built-in web, file, and computer tools.
February 2025
OpenAI launches Deep Research
OpenAI launched Deep Research on February 2, 2025, enabling longer web research tasks with cited reports.
December 2024
A General Language Assistant as a Laboratory for Alignment
Anthropic posted “A General Language Assistant as a Laboratory for Alignment.” A quick update for the record.
Discovering Language Model Behaviors with Model-Written Evaluations
Anthropic posted “Discovering Language Model Behaviors with Model-Written Evaluations.” A quick update for the record.
Language Models (Mostly) Know What They Know
Anthropic posted “Language Models (Mostly) Know What They Know.” A quick update for the record.
Measuring Faithfulness in Chain-of-Thought Reasoning
Anthropic posted “Measuring Faithfulness in Chain-of-Thought Reasoning.” A quick update for the record.
Measuring Progress on Scalable Oversight for Large Language Models
Anthropic posted “Measuring Progress on Scalable Oversight for Large Language Models.” A quick update for the record.