State of VLA Research at ICLR 2026
Research survey of Vision-Language-Action models at ICLR 2026. Covers VLA definitions, discrete diffusion, embodied reasoning.
Research survey of Vision-Language-Action models at ICLR 2026. Covers VLA definitions, discrete diffusion, embodied reasoning.
stereOS runs AI coding agents in sandboxed Linux VMs with credential injection. CLI tool (masterblaster) and pre-built mixtapes for rapid deployment.
Linux OS hardened for AI agents. Produces machine images with agent packages and restricted execution environments.
Commentary on AI generating mathematical proofs with hidden flaws or unverifiable complexity. Concerns about verification.
OpenCode tool for AI code review in CI/CD pipelines supporting non-GitHub/GitLab platforms with privacy-first design.
MCP-based local-first context engine providing persistent memory for AI agents (Claude, Codex, Cursor). Behavioral cloning via workflow pattern mining.
Semantic grep tool using Jina embeddings on MLX for Apple Silicon. Local-first, three search modes, no PyTorch dependency.
Research on Reality Alignment Index measuring AI system meaningfulness. Title and link only, insufficient content for evaluation.
Technical guide for deploying PyTorch models to Qualcomm NPUs via QNN SDK. Addresses integration challenges and provides developer experience improvements.
Praktor: Multi-agent orchestrator using Claude API with Docker isolation, Telegram interface, and Mission Control web UI. Go-based open source tool.
arXiv announcement about K-Search framework for LLM kernel generation via world models. Lacks technical details in provided content.
Shell command execution using natural language via zsh alias. Lightweight LLM application tool with practical developer utility.
Research evaluating memory structures in LLM agents. Title and link only, insufficient content for detailed assessment.
Tracecore benchmarking framework evaluates AI agents on deterministic ops tasks. Open-source tool with original research on agent reliability for infrastructure automation.
Usplus.ai platform for building autonomous AI agent teams to execute work. Open-source developer tool with functional implementation.
Itwillsync syncs terminal-based coding agents to mobile devices over LAN. Tool for agent interface/accessibility. Title only, minimal detail.
GIDE v1.0 AI code editor launch announcement. LLM application with limited detail provided.
Security analysis of OpenClaw skills revealing behavioral safety regressions. Demonstrates how well-written code can compromise agent safety despite passing static analysis.
Statistical analysis using Claude AI to prove Beale Ciphers hoax with Bayesian inference. Demonstrates LLM application to historical analysis with rigorous methodology.
Safari-CLI tool enables LLM agents to control Safari browser without MCP servers. Built for agentic development workflows.
Yaw terminal emulator combining terminal, SSH/database connections, and AI chat for Windows. Open source project.
Security audit guide for AI agent skills. Analysis of 31k+ skills found 485 critical and 1,718 high-severity vulnerabilities including prompt injection and supply chain attacks.
Personal AI assistant emphasizing security-first design for agent-based applications like calendar management and browser control.
OpenBrowserClaw browser-native AI assistant reimplementation. Runs entirely in browser tab with no infrastructure, supporting Claude API.
Game development using GitHub Codespaces and AI Copilot entirely in browser. Shipped Android game through Google Play with 100+ iterations.
Critique of LLM benchmark validity. Argues benchmarks lack signal due to test-set training and overfitting for social media hype.
Talentpluto voice AI agent platform matching GTM talent with startups using conversational AI and preference matching.
Challenge to build smallest transformer for 10-digit addition. Community competition comparing Claude Code (6,080 params) vs Codex (1,644 params).
Stellify: Structured code representation system enabling surgical AI-assisted development through queryable database storage instead of text files.
AppLaunchFlow uses AI to convert app screenshots into editable App Store creatives with design regeneration and video export.
EloPhanto: Open-source AI agent with 116 tools for video generation, email, animations using Remotion and physics simulation.
Benchmark study analyzing tool choices across 2,430 Claude Code runs. Finding: builds custom solutions over purchased tools in 85.3% of cases.
Open-source MIT-licensed context management tool reducing token costs 85% for LangGraph and OpenClaw agentic loops by optimizing conversation history re-reading.
27-line Claude Code persona based on Asimov's R. Daneel Olivaw to improve assistant reliability and reduce overconfidence.
Analysis of ChatGPT's writing style characteristics and how it leverages reader interpretation.
Study showing code review quality improves when asking Claude, Gemini, and Codex models to debate findings.
AI-native security firm uses AI tool to discover high-severity bug in Nethermind Ethereum software before exploitation.
Essay examining scalability limits of AI coherence in agentic systems and implications for future economy and AI engineering.
Batchling: Python library enabling AI developers to use provider-native batch APIs for 50% cost savings with standardized interface.
500-line Markdown-based agent orchestrator for Claude Code that outperforms multi-agent frameworks. Routes work across Claude, GPT, Gemini using filesystem IPC for inter-agent communication.
Cloud service offering managed Claude AI and hosting for web developers with fixed pricing and isolated testing environments.
Essay on open-source maintenance challenges and opportunities in early 2026, discussing AI's impact on software industry dynamics and sustainability.
Voice interface for Claude Code using live speech models for low-latency real-time conversations. Generic voice layer wraps any black-box agent without modifications.
Self-hosted LLM leaderboard ranking open-weight models across quality, speed, hardware requirements, and cost for enterprise deployment.
Depwire: Dependency graph and MCP tools providing AI coding assistants with codebase context to prevent hallucinated imports and missed dependencies.
pandō: CAD-like tool for code that reduces AI agent token usage and errors by treating code as structured data rather than text.
Local-first sovereign AI agent in Go/Rust with local LLM, data privacy, autonomous decision-making for personal digital life and financial management.
Open-source analytics platform for Claude Code sessions with archiving, cost tracking, multiplayer collaboration via GitHub/Google auth, and trend analysis.
GraphRAG agentic swarm system with 151k-node Neo4j graph using Claude 3.5 and Llama 3 for cross-domain reasoning on UN Sustainable Development Goals.
Open-source Claude Code clone in Rust with terminal UI for agentic workflows, tool execution (bash, file editing, web search), and session management.