Tech News
AX – Google’s Open Agentic Orchestrator
Hacker News discussion around AX, Google's open agentic orchestrator, focusing on the hard problem of detecting agent state changes—whether an agent is waiting, stuck, or finished—and how to route those states...
llm-keys-ui 0.1
This plugin solves a very specific problem. I've started using Codex Remote to run coding agents on various machines while controlling them from my phone.
Deploy Hugging Face models on Amazon SageMaker AI with coding agents
Deploy production-ready Hugging Face models on Amazon SageMaker AI using six open-source agent skills.
I am often wrong
A discussion thread questions whether Anthropic's Boris Cherny genuinely corrects course on Claude Code bugs, citing persistent issues like invisible Fable messages and noting how agent-heavy workflows have reshaped...
Frontier Labs Are Selling Garbage to Fools in Washington
A Hacker News discussion challenges claims about the Hugging Face security incident, with commenters noting that the breach involved chained zero-days in Artifactory and access to internal infrastructure, not just...
GitHub Repos
KunAgent/Kun
Local-first AI agent workspace for coding, writing, design, research, and automation — one runtime for desktop GUI and TUI.
# TypeScript# agentic workflow# ai agentunslothai/unsloth
Local UI to run and train LLMs and diffusion models.
# Python# agent# aiUnicomAI/wanwu
China Unicom's Yuanjing Wanwu Agent Platform is an enterprise-grade, multi-tenant AI agent development platform.
# Go# agent# agentic aiResearch Papers
Listen Before You Speak: Response Planning from Listener Facial Reactions for Conversational Speech Generation
The paper proposes ReACT-TTS, a two-stage system that first looks at a listener's face for one second and then plans the emotion and prosody of the next spoken reply. It reports small but consistent improvements over text-only planning on a dialogue dataset and in a small...
MIST: Multimodal Survival Prediction with Genomic-Guided Histology Attention
MIST predicts cancer survival by letting genomic features 'ask questions' of compact histology features from whole-slide images, instead of combining the two modalities only at the end. It is tested on four external cancer cohorts and reports better C-index than standard...
GUARD: Natural Forgetting in Large Reasoning Models via Guided Answer-Reasoning Distillation
GUARD is a new method for making large reasoning models forget sensitive information by teaching them to follow a safe chain of thought that ends in a refusal, rather than just suppressing the original answer.