Autonomous coding agent that runs each task in its own cloud sandbox preloaded with your repo — writing features, fixing bugs, running tests, and opening PRs. Reachable from ChatGPT web, a CLI, desktop apps, and IDEs (VS Code, JetBrains, Xcode).
Interleaves chain-of-thought reasoning with tool-using actions in one LLM loop: the model plans, queries a source like Wikipedia, then revises from results. Cuts hallucination versus reasoning-only prompting and beats trained agents on interactive tasks.
AI Agent Papers2024
Treats the interface between an LM agent and a computer as a design variable. A custom agent-computer interface (ACI) with concise file-edit, repo-navigation, and test commands plus compact feedback reaches 12.5% pass@1 on SWE-bench, 87.7% on HumanEvalFix.
Zirui Guo, Lianghao Xia +3University of Hong Kong, Beijing University of Posts and Telecommunications
A graph-based RAG framework pairing a knowledge graph with vector retrieval and a dual-level (low/high) query mode. New documents merge into the graph via set operations instead of triggering a rebuild, cutting the cost of keeping the index current.
AI Coding2025
Headless AI coding agent that runs a local HTTP server (OpenAPI 3.1) any client can drive — TUI, desktop, IDE plugins. Provider-agnostic: bring keys for any LLM, no vendor lock-in. Ships LSP-aware editing, plan/build agents, and shareable session links.
AI API2023
Run and manage open and community LLMs locally via a compact CLI and REST API—supports model import, Docker deployment, and official Python/JS SDKs for local inference, RAG, and dev workflows.
AI Train2023
Patches Hugging Face Transformers and TRL with hand-written Triton kernels to fine-tune LLMs on a single consumer GPU up to 30x faster with about 90% less memory. Does LoRA, QLoRA, and full fine-tuning across 500+ models, exporting to GGUF and Safetensors.
AI Others2015
Curated collection of AI courses, books, video lectures, papers and engineering guides for developers building generative models and agentic systems. Actively maintained with a weekly, evidence-backed curation workflow and an opinionated quality bar.
The 2017 paper that replaced recurrence with pure self-attention, making sequence models fully parallelizable — and, almost as a side effect, laying the architectural foundation for nearly every large language model that followed, from BERT to GPT.
AI Infra2017
Scales any Python or ML workload across CPUs and GPUs with a few decorators, instead of rewriting code for Spark or MPI. Bundles libraries for distributed training, hyperparameter tuning, RL, batch inference, and online model serving on one cluster.
Curates step-by-step, hands-on tutorials for reimplementing technologies from scratch—covering everything from OSs and compilers to neural networks, LLMs, and vision systems—so learners learn by rebuilding real systems across languages.