Skip to content
View najmulhasan-code's full-sized avatar
3 followers · 2 following

Achievements

Achievement: YOLOAchievement: QuickdrawAchievement: Pull Shark

Achievements

Achievement: YOLOAchievement: QuickdrawAchievement: Pull Shark

Block or report najmulhasan-code

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
najmulhasan-code/README.md

Najmul Hasan

Language Models · AI Alignment

My research interests center on language models and AI alignment. I am particularly interested in the design and training of language models, including how training choices shape their capabilities and behavior, and in developing models that are more capable, reliable, and aligned.

I completed a B.S. in Computer Science, with minors in Mathematics and Physics, at the University of North Carolina at Pembroke. I was an AI Safety Research Fellow at Algoverse, am currently participating in MIT AI Alignment's AI Safety Fundamentals program, and have completed BlueDot Impact's Technical AI Safety course.

Website · Google Scholar · LinkedIn · Twitter

Selected Publications

DPBench: Structural Determinants of Multi-Agent LLM Coordination Under Simultaneous Resource Contention
Najmul Hasan and Prashanth BusiReddyGari. Preprint, 2026. Code

Benchmarking Large Language Models for Zero-shot and Few-shot Phishing URL Detection
Najmul Hasan and Prashanth BusiReddyGari. LAW Workshop, NeurIPS 2025.

Honeypot Protocol
Najmul Hasan. AI Control Hackathon, Apart Research, 2026. Code

Open Source

SAGE
A Python framework in which language-model agents research, discuss, and synthesize answers through a structured workflow.

Sift
An autonomous IT ticket triage system that produces diagnoses, resolution steps, and escalation decisions.

Pinned Loading

  1. dpbench dpbench Public

    DPBench: A benchmark for evaluating coordination in multi-agent LLM systems under simultaneous resource contention.

    Python 2

  2. crc-screen crc-screen Public

    CRC-Screen: Certified DNA-Synthesis Hazard Screening Under Taxonomic Shift

    Python 1

  3. honeypot-protocol honeypot-protocol Public

    Honeypot Protocol

    Python 2

  4. splitcomp splitcomp Public

    Modeling when labs comply, evade, or split compute across jurisdictions and what drives each outcome

    Jupyter Notebook 1

  5. sage sage Public

    SAGE: Synchronized Agents for Generalized Expertise - A multi-agent framework where AI agents research, debate, and synthesize answers together

    Python 1

  6. sift sift Public

    Reads raw IT tickets. Returns structured resolution paths.

    Python 1