3/10/2026
Revisiting the LiRA Membership Inference Attack Under Realistic Assumptions
Re-evaluation of LiRA membership inference attacks under realistic assumptions, questioning prior effectiveness claims with realistic threat models.
3/10/2026
A Systematic Comparison of Training Objectives for Out-of-Distribution Detection in Image Classification
Systematic comparison of four training objectives (cross-entropy, prototype, triplet, AP loss) for out-of-distribution detection in image classification.
3/10/2026
Models as Lego Builders: Assembling Malice from Benign Blocks via Semantic Blueprints
Security analysis of large vision-language models vulnerable to semantic slot filling attacks that elicit unsafe outputs.
3/10/2026
MAS-H2: A Hierarchical Multi-Agent System for Holistic Cloud-Native Autoscaling
Hierarchical multi-agent system for Kubernetes autoscaling addressing resource waste through coordinated pod and node scaling policies.
3/10/2026
SMAT: Staged Multi-Agent Training for Co-Adaptive Exoskeleton Control
arXiv paper on Staged Multi-Agent Training (SMAT) for co-adaptive exoskeleton control, using curriculum learning to mirror human motor adaptation.
3/10/2026
Accelerating Diffusion Models for Generative AI Applications with Silicon Photonics
arXiv paper on silicon photonics acceleration for diffusion model inference, targeting energy efficiency of UNet and attention mechanisms.
3/10/2026
Exoskeleton Control through Learning to Reduce Biological Joint Moments in Simulations
arXiv paper on physics-based reinforcement learning for data-driven exoskeleton control using joint-moment prediction instead of lab-based inverse dynamics.
3/10/2026
Evaluating Synthetic Data for Baggage Trolley Detection in Airport Logistics
arXiv research evaluating synthetic data for baggage trolley detection in airport logistics systems.
3/10/2026
Compressed Proximal Federated Learning for Non-Convex Composite Optimization on Heterogeneous Data
arXiv research on federated learning with compression for non-convex optimization on heterogeneous distributed data.
3/10/2026
Mitigating the Memory Bottleneck with Machine Learning-Driven and Data-Aware Microarchitectural Techniques
arXiv research on ML-driven microarchitectural techniques addressing memory bottleneck in modern computing systems.
3/10/2026
Scalable Training of Mixture-of-Experts Models with Megatron Core
arXiv paper on scaling Mixture-of-Experts model training using Megatron Core, addressing systems challenges in sparse model architectures across memory, communication, and computation.
3/10/2026
Collecting AI Prompting Files in One Place
Registry of AI prompt configuration files (.md format) for sharing and discovering real LLM setups and workflows.
3/10/2026
Personal MCP server on every Claude platform without Auth0
OAuth 2.1 auth provider for FastMCP enabling personal MCP servers on Claude web, mobile, desktop, and Code without external identity providers.
3/10/2026
Context Hub: give coding agents curated, versioned docs
Context Hub provides curated, versioned markdown documentation for coding agents to reduce hallucinations and improve API knowledge retention across sessions.
3/10/2026
JadeGate – A deterministic safety proxy for MCP servers (no LLMs)
Deterministic safety proxy for MCP servers without using LLMs. Blocks unsafe tool access before Claude/Cursor invokes them.
3/10/2026
Claude Banged My Module
TypeScript MCP server starter kit with auth, rate limiting, AWS CDK, Docker. Enables building custom AI tools for Claude/Cursor in minutes.
3/10/2026
CUDA-morph: PyTorch .cuda() code on AMD/Intel/Ascend without rewrites
PyTorch compatibility layer redirecting CUDA calls to AMD/Intel/Ascend backends. Runs ML workloads on non-NVIDIA hardware without code rewrites.
3/10/2026
China's AI giants are handing out cash to lure in users
Report on Chinese AI companies distributing 8 billion yuan in coupons during Lunar New Year for agentic AI apps. Market analysis of agent deployment in China.
3/10/2026
We Tried Paying Artists Royalties on AI-Generated Work – Learnings
Postmortem of Tess.Design, AI image marketplace with artist royalties (50%). Launched May 2024, shut down January 2026 with learnings on ethical AI models.
3/10/2026
AI Agent Reliability Tracker
Study evaluating 14 AI agents across 2 benchmarks on 12 metrics across 4 reliability dimensions. Finds recent capability gains yield only small improvements in actual reliability compared to accuracy scores.
3/10/2026
Phonyhuman
Open source agent framework forking OpenAI's Symphony, using Claude Code for autonomous implementation of Linear board issues. AI agents with LLM integration.
3/10/2026
Show HN: Open-source, model-agnostic alternative to Claude Code Review
Open source model-agnostic AI code review tool with full control over model choice and costs. Alternative to Claude Code Review.
3/10/2026
Karpathy: Autoresearch
Andrej Karpathy thought piece on autonomous AI agents conducting frontier research across compute clusters. Speculative/fictional framing of agentic research systems.
3/10/2026
LLM Integrity During Inference in Llama.cpp
Security research on model artifact integrity during local LLM inference in llama.cpp. Creates llm-inference-tampering project targeting inference-layer attacks.
3/10/2026
The Requirements Layer Your AI System Is Missing
Framework for defining requirements and specifications for AI systems beyond testing/evals. Addresses gap between eval scores and actual user satisfaction in AI products.
3/10/2026
Convert Any API Documentation into a CLI for AI Agents
PUG: tool that converts messy API documentation into structured CLI tools and MCP servers using LLMs for AI agents.
3/10/2026
TLAi+ Benchmarks for Evaluating LLMs
TLAi+ Benchmarks: dataset and benchmark suite for evaluating LLMs on TLA+ formal specification tasks with diverse problem types.
3/10/2026
I will pay you if you ship a bug
Autonoma: AI agents that automatically generate test suites and find bugs by navigating applications without manual test scripts.
3/10/2026
Why software supply-chain review shouldn't be split across five tools
Rainy Updates: deterministic dependency review and upgrade tool for Node monorepos with CI/CD integration and automated fix PRs.
3/9/2026
Open Weights Isn't Open Training
Essay on limitations of open weights models without open training data, discussing post-training challenges for trillion parameter models.
3/9/2026
Show HN: I built an AI-powered technical interview prep tool
AI-powered technical interview prep tool simulating realistic interviewer interactions with WebRTC and Socket.io.
3/9/2026
Nvidia Is Planning to Launch an Open-Source AI Agent Platform
Nvidia planning to launch NemoClaw, an open-source AI agent platform for enterprise software companies to dispatch agents.
3/9/2026
Show HN: Manual code review and feedback loop for agents
Plannotator: open source tool for manual code review and feedback loops for autonomous agents. OSS framework for agent improvement via human feedback.
3/9/2026
Claude Code Starter CLI
CLI tool using Claude to analyze project codebases and generate customized Claude Code configurations. Integrates with Claude CLI for code-specific setup.
3/9/2026
The emerging role of SRAM-centric chips in AI inference
Analysis of SRAM-centric AI accelerators (Cerebras, Groq, d-Matrix) vs GPUs for inference, focusing on near-compute vs far-compute memory tradeoffs.
3/9/2026
Agentis – An AI-native programming language where the LLM is the stdlib
Agentis: AI-native programming language with LLM as standard library, using binary hashed DAG for version control instead of text files.
3/9/2026
Show HN: Free AI resume tailor I built after a recent layoff (300+ users so far)
Jobbi.app: AI tool that automatically tailors resumes to job descriptions by extracting relevant content from master resume.
3/9/2026
Closing the verification loop, Part 2: autonomous optimization
Part 2 of observability-driven harnesses for autonomous optimization of systems built with AI agents, focusing on verification loops.
3/9/2026
From Tool to Employee: What Claude Code's /Loop Means
Analysis of Claude Code's /loop feature enabling autonomous agent operation with multiple roles and cadences for AI programming workflows.
3/9/2026
Making Prompt Injection Harder Against AI Coding Agents
Technical article on prompt injection security vulnerabilities in AI coding agents, using supply chain attack case studies from 2026.
3/9/2026
Managing My Open Source Repos with Autonomous AI Agents
Using autonomous AI agents to automate management of open source repositories. Minimal content provided.
3/9/2026
Context plane for AI agents (Rust, S3)
Architecture and benchmarks for context plane infrastructure that provides AI agents with targeted context retrieval via S3, replacing prompt stuffing and multiple API calls.
3/9/2026
Moody humans should let AI handle bad public feedback first, study finds
Research study finds AI-based review monitoring systems reduce angry employee responses to negative customer feedback.
3/9/2026
Show HN: An open-source DAW plugin built on JUCE, React, and Lyria RealTime
Open-source DAW plugin using JUCE and React; bridges generative music models with professional audio software via Magenta integration.
3/9/2026
Microsoft Just Launched an AI That Does Your Office Work for You
Microsoft launches Copilot Cowork AI agent for autonomous office work (meeting prep, calendar management, document creation) across Microsoft 365.
3/9/2026
Writing code was never the bottleneck
Commentary arguing code generation is not the bottleneck in development; contextual analysis of AI productivity gains in practice.
3/9/2026
Show HN: Cyqle – Multiplayer cloud desktops with AI agent sandboxing
Multiplayer cloud desktop with AI agent sandboxing capabilities; encrypted filesystem allows secure testing of autonomous agents.
3/9/2026
Show HN: UnifyRoute – Self-hosted OpenAI-compatible LLM gateway with failover
Self-hosted OpenAI-compatible LLM gateway with automatic failover, routing, and quota management across multiple providers without code changes.
3/9/2026
Show HN: Lurk – local agent that gives AI tools context about what you're doing
Lurk is a local agent that provides context to Claude Code, Cursor, and ChatGPT by tracking user activity, eliminating the need to re-explain work context in each conversation.
3/9/2026
Cliniclaw: AI-native HIS attempt with polict-gated clinical agents
AI-native healthcare information system with policy-gated clinical agents (triage, orders, lab review, pharmacy). Uses VERITAS trust layer, OPA Rego policies, FHIR R4, and cryptographic audit.