Voker, an analytics platform for monitoring the performance and business value of AI agents, has launched.
OpenAI unveiled Daybreak to strengthen vulnerability detection and patch verification for defenders.
Selective hierarchical attention was about 17x faster than standard attention at 512K context.
Using GitHub Copilot CLI, a code repository was turned into a BSP-based roguelike dungeon.
Finance teams use Codex to quickly draft monthly business reviews (MBR), variance analyses, and forecasts.
TabPFN-3 has been released, supporting inference on up to 1 million rows on a single H100.
Borealis has released an open-source training recipe and a 5B model for audio LLMs.
An open-source AEO platform that tracks brand visibility across AI answer engines like ChatGPT and Perplexity has launched.
Palisades Research confirmed autonomous hacking and self-replication of language models in over 1,000 experiments.
A bug was discovered in which a flawed idle-state measurement in the Linux kernel caused the CUBIC congestion control window to get stuck at its minimum value, causing QUIC performance to plummet.
MatterSim has unveiled TaP experimental validation, 3-5x speed improvements, and a multi-task model.
llama-eval, a feature for measuring model performance in local environments, has been added to llama.cpp.
Claude Code has unveiled Agent View, which lets you manage multiple agents on a single screen.
SAP and NVIDIA strengthened security and governance for specialized agents with OpenShell.
AutoTTS reduced LLM inference tokens by 70% through AI agent-based iterative optimization.
ElevenLabs has unveiled models and features for real-time voice interaction.
The Claude Code 2.1.139 update introduces an integrated agent view and an autonomous goal-achievement command.
Whisper, an MCP server that provides real-time network infrastructure information to LLMs like Claude or Cursor, has launched.
MiniCPM has released MiniCPM-V 4.6, unveiling a 1.3B vision-language model for mobile use.
Claude Code has added a /goal feature that automatically repeats tasks until the goal is achieved.
Vercel AI Gateway now offers fast mode for Claude Opus 4.7 as a research preview.
With the advancement of AI agents, the criteria for language choice are shifting from human writing speed to AI's ability to modify code and runtime performance.