Tech News
Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows
Meta introduces Muse Glimmer, a 30B-parameter open-weight model optimized for always-on local agent workflows, with a focus on efficiency and portability.
Introducing Muse Glimmer
Meta are back in the open weights game! Muse Glimmer is a brand new 30B model under a clean Apache 2.0 license (a step up from the janky Llama licenses of old).
Build Low-Latency Multilingual Voice Agents: Open Weights & Full Deployment Control with NVIDIA Magpie TTS
NVIDIA released Magpie TTS, an open-weight text-to-speech model enabling low-latency, multilingual voice agents with full deployment control.
Run Local Agentic AI Workflows with Meta’s Muse Glimmer on NVIDIA
Meta returns to the open source ecosystem with the release of Muse Glimmer, a 30B open-weight dense model with a 120K+ context window built for local AI.
How nOps shipped FinOps agents 75% faster with Amazon Bedrock AgentCore
nOps rebuilt its Clara FinOps AI agent on Amazon Bedrock AgentCore, replacing a self-managed Amazon EKS stack running LangChain and LangGraph.
GitHub Repos
wshobson/agents
Multi-harness agentic plugin marketplace for Claude Code, Codex CLI, Cursor, OpenCode, GitHub Copilot, and Gemini CLI
# Python# agent skills# agentic ailobehub/lobehub
🤯 LobeHub is your Chief Agent Operator, organizing your agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.
# TypeScript# agent# agent collaborationdtyq/magic
Magicrew.
# TypeScript# agent# agiResearch Papers
TSPORec: Token Selection via Preference Optimization for LLM-Based Sequential Recommendation
This paper proposes TSPORec, a method that selects the most informative tokens from item descriptions to improve the efficiency and accuracy of LLM-based sequential recommendation systems.
Beyond the Capability Boundary: Zeroth-Order Optimization for Self-Evolving LLM Agents
This paper introduces a method that lets LLM agents improve beyond their current abilities by using zeroth-order optimization to adjust model parameters without needing labeled examples, then fine-tuning on the improved trajectories.
Test-Time Augmentation for LLMs: When Input Diversity Beats Output Diversity at Matched Compute
This paper compares two ways to spend extra compute during LLM inference: sampling multiple outputs (self-consistency) versus perturbing the input (test-time augmentation). It finds that semantic rephrasing of the input can convert compute into accuracy more efficiently than...