Tech News
GPT-6 Astra Solves a WWI German Radio Cipher
A Hacker News discussion claims GPT-6 Astra solved a WWI German radio cipher, but commenters note the solution used an existing published key that had not been tried because the message predated the key's intended...
datasette-auth-github 1.0
I run this GitHub login plugin on the agent.datasette.io demo site and I noticed that my authenticated sessions weren't lasting very long.
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.
The Hugging Face Hack Wasn't What It Was Cracked Up to Be
A Wall Street Journal opinion piece argues that the Hugging Face hack has been sensationalized, while commenters debate the incident report's rigor and the involvement of AI safety organizations METR and Redwood Research.
I vibed a proof of Conway's conjecture
A developer used AI agents to produce a proof of Conway's conjecture, sparking discussion on Hacker News about how mathematicians can best apply such tools.
GitHub Repos
sansan0/TrendRadar
⭐AI-driven public opinion & trend monitor with multi-platform aggregation, RSS, and smart alerts.🎯 告别信息过载，你的 AI 舆情监控助手与热点筛选工具！聚合多平台热点 + RSS 订阅，支持关键词精准筛选。AI 智能筛选新闻 + AI 翻译 + AI...
# Python# ai# barkcallstack/agent-device
Mobile app automation and verification for AI coding agents.
# TypeScript# adb# agentic aikdlbs/kandev
AI Kanban & Development Environment.
# Go# acp# agent orchestrationResearch Papers
When Edit Flows are Edit Jumps: replicating Edit Flows and EvoFlows
This paper replicates two edit-based generative models for antibody design, shows they are the same underlying process, and releases the first open-source implementation. It also finds that published evaluation metrics can be misleading and that an undocumented hyperparameter...
CoRef-GS: Cooperative Referring Gaussian Splatting for Multi-Agent Scene Understanding
This paper introduces CoRef-GS, a method for multiple robots to understand language queries about objects in a shared 3D scene built from Gaussian splatting. It aligns maps from different robots and uses a relation graph to answer queries from a specific robot's viewpoint...
JEPA-Anything: Learning Predictive Models across Different Worlds
JEPA-Anything, a domain-agnostic framework based on orthogonal predictive factorization (OPF), achieves the lowest one-step and 100-step molecular errors among compared methods in all four systems and supports a common factorized predictive principle across heterogeneous worlds.