Skip to content
View rapturt9's full-sized avatar
8 followers · 3 following

Achievements

Achievement: YOLOAchievement: Pull Sharkx2

Achievements

Achievement: YOLOAchievement: Pull Sharkx2

Block or report rapturt9

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 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
rapturt9/README.md

Hi, I'm Ram Potham 👋

I'm an AI Safety Researcher at the MIT Algorithmic Alignment Lab

About Me

I'm focused on ensuring the development of advanced AI leads to a safe and prosperous future. My perspective is shaped by my prior experience as the founder of a VC-backed startup where I built autonomous AI agents. This gave me a firsthand understanding of the rapid progress and potential risks in AI, motivating me to pivot my career to focus on them. My research focuses on mitigating existential risk from AI.

🚀 Featured Publications

  • Evaluating LLM Agent Adherence to Hierarchical Safety Principles

    • Description: A lightweight benchmark for evaluating an LLM agent's ability to uphold a high-level safety principle when faced with conflicting instructions.
    • Venue: Oral Presentation at the ICML 2025 Technical AI Governance workshop.
    • ➡️ Read the paper on arXiv (2506.02357)

  • MAEBE: Multi-Agent Emergent Behavior Framework

    • Description: A framework for analyzing emergent behaviors in multi-agent systems, focusing on safety and alignment in complex AI environments.
    • Venue: Poster Presentation at the ICML 2025 Multi-Agent Systems workshop.
    • ➡️ Read the paper on arXiv (2506.03053)

💻 Tech Stack & Skills

  • AI Safety Concepts: Empirical Alignment, Safety Evaluations, Robustness Testing, Guardrails, Safety Cases
  • Languages & Frameworks: Python, PyTorch, TensorFlow
  • ML & Engineering: Multi-Agent Systems, LLM Fine-Tuning, AWS, Cloud Architecture

Pinned Loading

  1. pik-gane/satisfia pik-gane/satisfia Public

    Satisficing-based Intelligent Agents

    Python 5 2

  2. wisdom_agents wisdom_agents Public

    How multiple agents moral responses influence each other

    Jupyter Notebook 1 2

  3. SafetyAdherenceBenchmark SafetyAdherenceBenchmark Public

    LLM Agent Principle Adherence Benchmark

    Jupyter Notebook 1

  4. Algorithmic-Alignment-Lab/character-training Algorithmic-Alignment-Lab/character-training Public

    Python