Skip to main content
← All Tags

Agentic AI

195 articles in this category (Page 4 of 9)

AI NewsAgentic AIAI Agents

Building Advanced Cybersecurity AI Agents with the CAI Framework

The CAI framework enables modular cybersecurity agents with multi-agent handoffs, custom function tools, and heuristic guardrails for automated threat intelligence.

Read more
AI NewsAgentic AIMachine Learning

Amazon Researchers Release A-Evolve: An Automated Evolution Framework for AI Agents

A-Evolve automates AI agent development, achieving a 79.4% top score on the MCP-Atlas benchmark by replacing manual prompt tuning with automated state mutation.

Read more
AI NewsAgentic AIAI Infrastructure

Agent-Infra AIO Sandbox: A Unified Execution Layer for AI Agents

Agent-Infra releases AIO Sandbox, an open-source runtime integrating Chromium, Python, and Node.js into a unified filesystem for agentic AI.

Read more
AI NewsAgentic AIAI Agents

Exploring nanobot: A Lightweight 4,000-Line Python Framework for AI Agent Pipelines

Learn to build full agent capabilities using nanobot, an ultra-lightweight 4,000-line Python framework for memory, tools, and subagent delegation.

Read more
AI NewsAgentic AISEO

Google-Agent vs Googlebot: Understanding the Technical Shift to User-Triggered AI Fetching

Google defines Google-Agent as a user-triggered fetcher that bypasses robots.txt to retrieve real-time content for AI tools, distinct from Googlebot's autonomous crawling.

Read more
AI NewsAgentic AIKnowledge Graphs

Implementing IWE's AI-Powered Knowledge Graph for Agentic RAG

Michal Sutter demonstrates IWE, a Rust-powered system that transforms markdown notes into a navigable knowledge graph for multi-hop AI reasoning across 8 developer documents.

Read more
AI NewsAgentic AIAI Agents

Self-Evolving AI Agents: JiuwenClaw Launches with Autonomous Skill Optimization

OpenJiuwen Community releases JiuwenClaw, a self-evolving AI agent that uses hierarchical memory and local environment takeover to solve task execution bottlenecks.

Read more
AI NewsAgentic AIAI Agents

openJiuwen Releases JiuwenClaw: A Self-Evolving AI Agent for Execution-Centric Task Management

openJiuwen Community launches JiuwenClaw, a self-evolving AI agent featuring a hierarchical memory system and autonomous skill refinement for complex task execution.

Read more
AI NewsLarge Language ModelAgentic AI

Implementing Qwen3.5 Claude-Style Reasoning with GGUF and 4-Bit Quantization

Run a 27B Qwen3.5 distilled reasoning model on a single GPU using 4-bit quantization and GGUF for optimized inference under 16.5 GB of VRAM.

Read more
AI NewsAgentic AILarge Language Model

Google Releases Gemini 3.1 Flash Live: Real-Time Multimodal Voice for AI Agents

Google launches Gemini 3.1 Flash Live, a low-latency multimodal model achieving 90.8% on ComplexFuncBench Audio for real-time voice-first AI agents.

Read more
AI NewsAgentic AIAI Agents

Building Vision-Guided Web Agents with MolmoWeb-4B and Multimodal Reasoning

MolmoWeb-8B achieves a 78.2% pass@1 rate on WebVoyager by interacting with websites via screenshots without DOM parsing, using a 2.2M screenshot QA dataset.

Read more
AI NewsAgentic AIMachine Learning

NVIDIA AI Introduces PivotRL: Efficient Agentic Training with 4x Fewer Rollouts

NVIDIA’s PivotRL framework achieves high agentic accuracy using 4x fewer rollout turns and training 5.5x faster than end-to-end RL.

Read more
AI NewsAgentic AITutorials

OpenSpace: A Self-Evolving Skill Engine for Autonomous AI Agents

HKUDS introduces OpenSpace, a self-evolving engine achieving 46% token reduction and 4.2x income improvement in real-world professional tasks.

Read more
AI NewsAgentic AIModel Context Protocol (MCP)

Automating Google Colab with AI Agents: A Guide to colab-mcp and FastMCP

Design a production-ready AI agent that automates Google Colab workflows using the open-source colab-mcp server and FastMCP framework.

Read more
AI NewsAgentic AILarge Language Model

Luma Labs Uni-1: Bridging the Intent Gap with Autoregressive Reasoning Transformers

Luma Labs Uni-1 utilizes a decoder-only autoregressive transformer to reason through spatial logic before generation, outperforming Flux Max on RISEBench at $0.10 per image.

Read more
AI NewsAgentic AIAI Infrastructure

Meta AI Hyperagents: Achieving Recursive Self-Improvement via Metacognitive Self-Modification

Meta AI's DGM-H hyperagents achieve 0.710 performance in paper reviews by rewriting their own improvement logic without manual intervention.

Read more
AI NewsAgentic AIMachine Learning

LeWorldModel: Yann LeCun’s End-to-End JEPA for Pixel-Based Predictive Modeling

LeWM achieves 48x faster planning than DINO-WM using a stable end-to-end JEPA architecture with only two loss terms and SIGReg regularization.

Read more
AI NewsAI InfrastructureAgentic AI

GitAgent: A Universal Open-Source Format for Framework-Agnostic AI Agents

GitAgent introduces an open-source CLI tool to decouple AI agent logic from frameworks like LangChain and AutoGen using a Git-native architecture for better portability.

Read more
AI NewsAgentic AIAI Agents

Implementing Multi-Agent Swarm Orchestration with ClawTeam and OpenAI Function Calling

Michal Sutter implements HKUDS's ClawTeam framework using OpenAI function calling to orchestrate agents and resolve task dependencies in a Colab environment.

Read more
AI NewsAgentic AILarge Language Model

NVIDIA Nemotron-Cascade 2: High-Density 30B MoE with Gold Medal Reasoning

NVIDIA’s Nemotron-Cascade 2 is a 30B MoE model with 3B active parameters achieving Gold Medal-level results in IMO and IOI reasoning benchmarks.

Read more
AI NewsAgentic AISoftware Engineering

Google Colab MCP Server: Programmatic AI Agent Access to GPU Cloud Runtimes

Google releases the open-source Colab MCP Server, enabling AI agents to autonomously execute Python code and manage cloud-hosted GPU runtimes via the Model Context Protocol.

Read more
AI NewsAgentic AIOpen Source

LlamaIndex LiteParse: TypeScript-Native Spatial PDF Parsing for AI Agents

LlamaIndex launches LiteParse, a local-first, TypeScript-native library for spatial PDF parsing designed to optimize data ingestion for AI agents.

Read more
AI NewsAgentic AIAI Infrastructure

NVIDIA Open-Sources OpenShell: Secure Sandboxed Runtime for AI Agents

NVIDIA released OpenShell under Apache 2.0, a secure runtime providing kernel-level sandboxing and L7 policy enforcement for autonomous AI agents.

Read more
AI NewsAgentic AILarge Language Model

ServiceNow Research Launches EnterpriseOps-Gym to Benchmark LLM Agentic Planning

ServiceNow Research's EnterpriseOps-Gym reveals that even top LLMs like Claude Opus 4.5 fail to exceed a 37.4% success rate in enterprise planning tasks.

Read more