#Autonomous
Curated coverage on «Autonomous»: releases, research, benchmarks, open-source.
Tencent released Hy3—an open MoE model with 295 billion parameters and 256K context
Tencent presented Hy3—a model with 295 billion parameters, 21 billion active per token, Apache 2.0 license and a 78.0 score on SWE-Bench Verified.
Neuroanalytic 2.0 in Yandex DataLens: overview of an AI agent for data analysis
Yandex introduced Neuroanalytic 2.0 — an updated AI assistant in the BI system DataLens, which now works as a full-fledged AI agent rather than just a data…
How to evaluate AI agent libraries: the example of transformers and traces over answers
Developers of AI agent libraries are moving from checking the correctness of the final answer to analyzing the agent's path itself — traces, tokens, time and…
Autonomous AI agents create critical data loss risk in DevOps
NVIDIA Alpamayo: Closed-Loop Model Training for Autonomous Vehicles
Autonomous AI agents run 50x longer than search: Harvard and Perplexity research
Memory OS: Open 6-Layer Memory Stack for Hermes Agent
AWS adds secrets management from Secrets Manager to Bedrock AgentCore
AWS Secrets Manager is now integrated with Bedrock AgentCore Identity, allowing developers to use pre-configured secrets and manage their lifecycle.
Google DeepMind funds research on risks of interaction between millions of AI agents
Google DeepMind issued grants to study potential dangers when millions of different AI agents interact with each other online without human oversight.
C3 AI deploys agents for predictive maintenance at Shell
Xebia: AI Agents Fail Without Proper Data Foundation
Nous Research presented Hermes Agent profile builder
GitHub Agentic Workflows: Automate Documentation After Commit
AI Models Outpace Security Tests: Regulators Lose Visibility into Capabilities
Advanced AI models have begun to outpace benchmarks designed to measure their cybersecurity, leaving regulators and security experts partially blind to the real…
AI Programmer Assistants Trigger Sophos Alerts
Routine actions by AI programmer agents trigger false positives in Sophos antivirus behavioral detectors, but this is not a software limitation.