Skip to content
View ai-nikolai's full-sized avatar
🌙
Building Stuff
🌙
Building Stuff
39 followers · 92 following

Achievements

Achievement: YOLOAchievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Starstruckx2Achievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Achievements

Achievement: YOLOAchievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Starstruckx2Achievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Block or report ai-nikolai

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ai-nikolai/README.md

👋 Hi there!

I am a CS PhD student in LLMs & LLM agents at Imperial College London with Marek Rei. That's my latest work: StateAct, which outperforms ReAct by ~10%-30%.

Checkout my website: nikolairozanov.com

I work on:

  • 🧠 LLM Reasoning
  • 🛠️ LLMs + Tools
  • 🤖 LLM agents

🧰 Toolkits

Python C++ CUDA JavaScript LateX

PyTorch TensorFlow HuggingFace Unsloth vllm

NumPy Pandas SciPy matplotlib SpaCy scikit-learn Jupyter Anaconda Git

Linux macOS

Pinned Loading

  1. StateAct StateAct Public

    StateAct - REALM25 @ ACL25 - Official Paper Repo (SOTA LLM Agent)

    Python 9

  2. simple_vllm simple_vllm Public

    A simple wrapper around vllm to allow for quick inference and OpenAI compatible API.

    Python 1

  3. textcraft textcraft Public

    [Pypi Repo] for Textcraft Environment

    Python 1

  4. LLamp LLamp Public

    Larage Language Model Planning (LLAMP)

    Jupyter Notebook 2

  5. Wluper/matilda Wluper/matilda Public

    MATILDA: Multi-AnnoTator multi-language Interactive Lightweight Dialogue Annotator

    JavaScript 149 28

  6. Wluper/Retrograph Wluper/Retrograph Public

    Common Sense or World Knowledge? Investigating Adapter-Based Knowledge Injection into Pretrained Transformers

    Python 20 5