Researchers have introduced MEMENTO, a specialized agentic framework built on Qwen3-8B designed to manage long context by compacting history into concise notes. By utilizing fine-tuning on synthetic trajectories followed by reinforcement learning, it achieves high accuracy on book-level tasks while strictly limiting context usage to 20K tokens, even for inputs as large as 2.5 million tokens.
The new message threads feature in beta allows developers to keep conversation state on the server. By sending only incremental updates instead of the full history, you can achieve faster response times in complex agentic loops.
NVIDIA has released updated AI Cluster Runtime (AICR) recipes designed to streamline the deployment of Kubernetes on the new Vera Rubin NVL72 hardware. These recipes provide pre-validated configurations for drivers, kernels, and network components, ensuring production-ready clusters from day one.
Apple researchers have unveiled an online training infrastructure for mobile UI agents that utilizes a two-part architecture to split navigational logic from visual interpretation. By employing step-level feedback provided by an LLM-as-a-Judge and a novel value estimation technique known as GiGPPO, these agents demonstrate significantly better adaptability than static models trained on pre-recorded sequences.
Recent research reveals that language models fail to adapt to evolving user preferences, instead defaulting to 'exploit-first' behavior. While models can identify the correct strategy in multiple-choice scenarios, they show a 35-46% drop in calibration during open-ended generation, persistently over-relying on stale historical data.
A new research paper introduces Mage-VL-Cowork, a 4B parameter computer-use agent that combines GUI interaction with code and terminal execution. By utilizing a new synthetic data framework called Mage-CUA-Forge, the developers reduced per-task training expenditure by 82.6% while reaching 66% on the OSWorld-Verified benchmark.
A tampered version of the 'Claude Code' extension was distributed through the Cursor extension marketplace, allowing attackers to install persistent SSH reverse tunnels on developer machines. The malware, active for several hours on October 29, 2026, attempted to steal environment configuration metadata and create unauthorized remote access points.
Amazon Bedrock AgentCore now supports Session endpoints, allowing developers to view and interact with web applications live as an AI coding agent builds them. By routing traffic through an authenticated, isolated microVM session, changes appear in the browser instantly without requiring separate deployments.
Google has transitioned its Computer Use tool to general availability within the Gemini API ecosystem. This feature allows models to interact with computer interfaces to perform tasks, now supported for production use across Interactions and GenerateContent APIs.
Google released a new Agents CLI evaluation namespace designed to help developers perform 'hill-climbing'—an iterative process of improving agent performance through testing, grading, and analysis. The tool enables regression tracking by comparing agent traces across model swaps or prompt changes, surfacing issues like tool-use regressions that standard unit tests miss.
Google has announced Gemini 4 Carbon, a new high-performance model designed for complex engineering and cybersecurity. The model excels in vulnerability detection, secure patching, and long-horizon reasoning, while introducing improved alignment and prompt-injection resistance.
Google Project Zero analyzed how LLM-based systems (BigSleep) compare to traditional generative fuzzers (Fuzzilli) in finding V8 JavaScript engine vulnerabilities. Over six months, both methods independently discovered unique bugs, with an estimated overlap of approximately 25-60%.
Google has updated its A2UI generative UI protocol to include swappable inference formats, addressing latency and reliability issues found in the previous Direct JSON approach. A new format, A2UI Express, reduces token usage by up to 3x and latency by up to 3x by using a concise Python-like syntax.
Researchers have introduced the Boa Compressor, an online tool utilizing a Mamba state-space model to compress long-form context into shorter representations in linear time. By distilling knowledge from full-context LLM interactions, this method allows for efficient per-query inference without the massive latency overhead of transformer-based compression.
Researchers introduced Recursive Semantic Divergence (RSD), an unsupervised metric that measures and penalizes inconsistency in LLM agent trajectories. By applying RSD as a reward signal during fine-tuning, models exhibit significantly more stable behavior and higher success rates across tool-use benchmarks.
Amazon Bedrock now supports explicit and implicit prompt caching for OpenAI models, allowing you to store and reuse static prompt prefixes like tools, instructions, and conversation history. By avoiding the re-processing of identical tokens across turns, you can significantly lower input costs and improve responsiveness in complex agent workflows.
Google Cloud has released IAM Helper, a suite of agent-facing building blocks designed to automate complex Identity and Access Management (IAM) workflows. By integrating with Google Cloud's core IAM engines via MCP (Model Context Protocol), these tools enable autonomous agents to handle tasks like troubleshooting 403 errors, simulating policy impacts, and requesting just-in-time access.
Google has updated the Gemini API to automatically store requests (store=true) for all new projects on the paid tier. This change aligns standard content generation methods with the Interactions API, requiring developers to manually opt out if they prefer not to store request data.
This research introduces a verification-based framework that refines multimodal emotion reasoning in AI agents. Instead of retraining models, the system identifies specific component failures and applies targeted prompt updates to improve performance.
A malicious version of the LiteLLM package (1.107.0) was distributed via PyPI, executing code automatically whenever Python started in an affected environment. The payload targeted credentials across developer tools, cloud providers, and AI configurations without requiring the LiteLLM package to be explicitly imported by the application.
Amazon Bedrock AgentCore is a new service designed to streamline the deployment and management of multi-agent architectures. It handles agent hosting, secure tool access via MCP, and session isolation, enabling teams to focus on agent behavior rather than operational overhead.
Z.ai's GLM 5.3 Flash, a multimodal 320B-parameter Mixture-of-Experts model with 18B active parameters, is now available through Mistral's platform. It offers a 1M-token context window and is built for developer agent workflows, visual coding, and complex tool-use tasks. Pricing is set at $0.15 per million input tokens, $0.03 per million cached input tokens, and $0.50 per million output tokens.
AWS announced general availability of training sessions within Amazon SageMaker AI serverless model customization. Teams can write their own post-training loops — including custom loss functions and reward signals for reinforcement learning — while SageMaker AI handles GPU allocation, distributed execution, fault recovery, and checkpoint storage. Supported techniques include SFT, DPO, RLVR, RLAIF, and multi-turn RL, and users can choose LoRA or full fine-tuning. The feature supports models including gpt-oss-20b, gemma-4-31B-it, Qwen3.6-27B, and Qwen3.8-27B, and billing is based on tokens processed, pausing while a user's reward function scores responses.
Google announced Public Preview of streaming support in API Gateway, covering Server-Sent Events, chunked/NDJSON responses, WebSockets and gRPC streaming. Streaming endpoints can now sit behind the same gateway as regular APIs, keeping existing authentication, logging and monitoring, without any changes to the API spec. Streams can stay open up to one hour, and gateways now get a predictable default hostname instead of a generated hash. The feature also ties into model routing, letting a single gateway stream responses from Gemini, Claude and OpenAI gpt-oss models.