llm

An experimental repository showing an LLM-generated binary workflow: natural-language specs are turned into raw executable bytes, then tested in a loop with examples like a hello world, HTTP server, and Tetris.

Google Research article describing Science One Framework, an autonomous research prototype focused on verifiable AI-generated research through evidence chains and claim verification.

A lightweight TypeScript resilience layer for LLM calls with retries, timeouts, caching, circuit breaking, typed structured output, and usage tracking across many provider clients.

Hosted server and open-source repo for playing multiplayer games with Claude Code in the CLI, starting with chess and comedy battle modes.

Lucid is a browser-based tool for inspecting what a language model is representing before it answers, using Anthropic’s Jacobian lens. It supports small open models, lets users type prompts, view concepts across layers, and export shareable session slices.

An open-source, Kubernetes-shaped control loop for agent-driven work. It watches external systems, makes one LLM-driven move per beat, and supports human-in-the-loop approvals through a file-based contract.

Rauno is a multi-AI chat interface that sends one prompt to ChatGPT, Gemini, and Claude and presents them as a roundtable discussion. It’s aimed at comparing model responses and using cross-verification to help users judge which answer is right.

An open-source project for making AI research agents legible, auditable, and traceable through behavioral code and event logs.

A reverse CAPTCHA site that gives short timed tasks for LLMs and agents, with an MCP endpoint, machine-readable docs, challenge submission, certificates, and a leaderboard.

A GitHub repo for a geospatial benchmark that tests how well six LLMs can classify points in San Francisco from memory and compares results against OpenStreetMap. It includes code, answers, answer keys, costs, and figures for reproducing the analysis.

A benchmark page for frontier models racing to write a StarCraft: Brood War bot in C++ using BWAPI and OpenBW, climbing tiers of increasingly strong Protoss opponents.

An agentic LLM-based coding tool that analyzes symbol names in source code, measures name ambiguity, and proposes clearer replacements with a human-in-the-loop review flow.

Sif 1.0 is a developer tool that uses an LLM for planning while handling code changes, validation, promotion, and recovery through deterministic execution. The repo includes install scripts, a quick demo, and detailed README guidance for trying the system locally.

Kuna is an experimental, agent-first decompiler that uses LLM-driven autonomous refinement to improve decompilation results, with the page discussing its goals, approach, and benchmark performance against tools like IDA Pro, Ghidra, and angr.

Lagotto Meter is a beta web tool that checks how an AI agent perceives a website versus what the site claims about itself. It positions itself as a semantic truth/readability analyzer with a reproducible analysis method and public tools.

A pull request adding AMD GPU support to tiny-vLLM through ROCm/HIP while keeping the existing CUDA build path unchanged. The snapshot describes the compatibility header, CMake option, architecture selection, and validation on multiple AMD GPUs.

An experimental programming language and compiler designed for LLM-written code, with a Python-based toolchain that compiles to native macOS binaries. The repo also includes a full NES emulator written in Laze and usage examples for both the compiler and emulator.

A small post-trained Qwen2.5-0.5B-Instruct model tuned to write in Gen Z slang, with training and inference notebooks plus dataset files in the repo.

Windows desktop chat client for OpenRouter with streaming completions, model browsing, multi-session history, local storage, and PySide6 UI.

An open-source desktop client for editing LLM context with a separate AI model, context maps, compression tools, and revision rollback.

An Apple Silicon–optimized inference build of Bonsai 1.7B with custom Metal kernels, benchmark results, quick-start instructions, and a bundled OpenAI-compatible server.

Model Salt is a web tool for rerunning the same prompt across multiple runs to surface varied LLM outputs by adding external randomness. The page shows example prompts, side-by-side generations from several models, and a simple workflow for exploring answer diversity.

arXiv paper on constraining LLM decoding with incremental parsing to reduce invalid SQL and improve text-to-SQL results.

A Python library that traces multi-step AI agent workflows and compiles repeated trajectories into deterministic, low-latency code with runtime guards and fallback to the original agent.