An open‑source, student‑driven research laboratory focused on LLMs and Computer Vision. Building the mathematical foundation for tomorrow's AI.
Our commitment to innovation, transparency, and high‑quality research output.
We believe knowledge belongs to everyone.
All research, datasets, and models produced by ITM AIR Lab are released openly to the community.
We don't just use API wrappers.
Our focus is on fundamental development in Large Language Models (LLMs) and Computer Vision Models (CVMs).
Building trust through visibility.
Daily updates on our website regarding research progress, published papers, and lab news.
A structured approach to tackling complex AI problems.
Explore our suite of open-source tools and visualizers.
From a student initiative to a state-level AI hub.
Complete autonomy. Students handle research, development, dataset creation, and website updates. A temporary committee of HODs provides administrative support and approvals without restricting research freedom.
Formal governance involving Senior Faculty, External AI Researchers, and Industry Advisors to oversee long-term strategy, funding, and ethical standards.
To ensure academic credibility and public trust, we maintain a daily-updated digital presence.
Understanding the force that is reshaping our world and why we must lead it.
Rule-Based Systems (1950s-90s): Logic and if-then statements. Rigid and unable to learn.
Deep Learning (2010s): Neural networks that learn from vast data, mimicking the human brain.
Generative AI (Today): Models that create, reason, and code. The era of LLMs and Foundation Models.
Healthcare: Diagnosing diseases earlier with higher accuracy than human doctors.
Automation: Handling repetitive tasks, freeing humans for creative problem solving.
Creativity: Democratizing art and coding, allowing anyone to build software.
As AI becomes more powerful, it becomes more opaque. Most advanced models are proprietary "black boxes" owned by a few corporations.
This creates a dependency risk and limits scientific understanding.
The world needs more than just API consumers. We need scientists
who understand the mathematics of attention mechanisms, loss
landscapes, and alignment.
At ITM AIR Lab, we don't just use AI; we build it. We bridge the
gap by empowering students to conduct
fundamental, open-source research
that is transparent, ethical, and accessible.
Latest research, technical breakdowns, and thoughts from the lab.
An in-depth analysis of how self-attention mechanisms are evolving beyond standard LLM architectures to create more efficient, edge-deployable models.
How our latest experiments with semantic segmentation are helping to identify early-stage anomalies in X-ray and MRI scans with 98% accuracy.
We explore the 'black box' problem and propose a new framework for transparent governance in student-led AI research initiatives.
The 3rd year students of the AIML branch leading the lab's research and development efforts.
Research Head
AIML (3rd Year)
Data Head
AIML (3rd Year)
Developer Head
AIML (3rd Year)
Faculty members providing academic guidance, institutional support, and strategic oversight.
HOD of CSE
Strategic Oversight
HOD of ME
Institutional Support
Head of CSE-Specialization
Academic Liaison
x⁴ - πx² + (e-1)y² = 0
Our logo is defined by this specific equation, representing the rigorous mathematical foundation of our models, the balance in our intelligence systems, and the infinite pursuit of knowledge.
Join the Initiative