Skip to content
View Axelgustavlindstrom's full-sized avatar
  • Italy

Block or report Axelgustavlindstrom

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.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

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

Report abuse
Axelgustavlindstrom/README.md

Axel Gustav Lindström avatar

Axel Gustav Lindström

GitHub Source

Stockholm, Sweden

About Me

I am a researcher and engineer focused on making AI systems more reliable, aligned, and capable through feedback-driven learning and structured tool use.

My work centers on Reinforcement Learning from AI Feedback (RLAIF), Model Context Protocol (MCP), test-time training (TTT), and building robust agentic systems. I am particularly interested in how models can improve during inference, how they can safely interact with external tools via standardized protocols, and how automated feedback loops can be used to align behavior without relying solely on human annotation.

I value clean interfaces, reproducible experiments, and minimal abstractions that scale.

Research Interests

Focus Area Description
RLAIF Using AI-generated feedback to steer policy improvement and preference learning
MCP Standardizing how agents and models interact with tools and external systems
TTT Adapting or fine-tuning behavior at inference time for improved performance
Agentic Systems Building reliable, tool-using agents that plan, execute, and recover

Tech Stack

Python PyTorch Jupyter Git Linux GitHub Markdown

Tools

  • Training & evaluation: PyTorch, Jupyter
  • Version control: Git
  • Environment: Linux
  • Writing & docs: Markdown

Skills

Skill Area
Reinforcement Learning RLAIF, RLHF behavior
Model Context Protocol Tool interfaces & agent integration
Test-Time Training Inference-time adaptation
Agentic Systems Tool use, planning, execution
Python Core development & research

Languages

  • Python — primary language for research and engineering

Popular repositories Loading

  1. Axelgustavlindstrom Axelgustavlindstrom Public

    GitHub Profile README for Axel Gustav Lindström — researcher and engineer focused on RLAIF, MCP, TTT, and agentic systems.

  2. eventbroker-lite eventbroker-lite Public

    A lightweight typed in-process event broker with topic-based pub/sub, middleware hooks, optional SQLite persistence, and a new package entry point.

    Python

  3. cmclean cmclean Public

    A comment-structure linter for source trees. It scans source files and flags commented debris: TODOs/FIXMEs/HACKs, author plaques, suspicious one-liners, excessively long comments, and blocker-styl…

    Python

  4. gitremind gitremind Public

    Local git audit tool that flags commits whose touched file set no longer appears on the branch tip.

    Python

  5. commitlint-local commitlint-local Public

    Repo-local Conventional Commits linter and pre-commit hook with configurable rules.

    Python

  6. pypinventory pypinventory Public

    Scan Python project dependencies and generate actionable inventory reports.

    Python