Machine learning researcher · IISER Bhopal

Learning representations that remain reliable beyond the familiar.

I am Mohammed Kaif Kalagond, a B.S. Data Science student at IISER Bhopal (CGPA 9.18/10). I study robust learning, fairness, representation geometry, and world models.

Research direction

I build methods that make intelligent systems more stable, more fair, and more dependable when the world changes.

Selected research

Questions I am pursuing

I work across representation learning, robustness, and fairness - with an emphasis on mechanisms inside the model that can be measured, regulated, and tested.

01 / Thesis2026 - present

Stable long-horizon prediction in world models

Investigating manifold-based latent representations for world models, focusing on representation drift and error accumulation during recursive prediction, and how latent geometry and dynamics affect the stability of long-horizon predictions.

World modelsLatent geometryDynamics
02 / Robust learningICLR 2026

Regulating internal alignment flows

Co-developed Alignment-Gated Suppression, a regulariser that uses class-conditional, confidence-weighted internal alignment signals to suppress reliance on spurious features without group labels. Evaluated on Waterbirds, CelebA, BAR and COCO with under 5% training overhead.

Spurious correlationsWorst-group robustness
03 / Fairness2025 - present

Localising and removing bias in vision models

Studying where attribute bias lives across layers and representation subspaces, and building adversarial and representation regularisers that mitigate multiple biases while preserving accuracy - from skin-tone-fair dermatology to multi-attribute debiasing.

DebiasingMedical imagingRepresentation analysis

Publications & manuscripts

Selected work

2026

Manuscript · Submitted to Nature Aging

Chronological-age prediction, long-term stability and prospective host relevance are separable properties of gut microbiome ageing measures

Radhika Khatri, Adinath Dukre, Anastasia Godneva, Mohammedkaif Rafiq, Kabir Suhel, Eran Segal, and Imran Razzak.

2027

Manuscript · Submitted to ICLR 2027

Bias-Enhanced Attention for Mitigating Spurious Correlations

Under review.

2027

Manuscript · Submitted to ICLR 2027

Mask-Guided Attention for Debiasing Neural Networks

Under review.

2026

Manuscript · Submitted to WSDM 2026

Bias-Free Skin Lesion Classification via Robust Representation Learning

Under review.

2025

Project · Research prototype

Multi-Agent Medical Reasoning with LLaMA

A pipeline of seven LLaMA models assigned to domain identification, expert analysis, and response summarisation for structured medical reasoning.

Experience & service

In the lab and community

2026 - present

B.S. Thesis Student · VisDom Lab, IISER Bhopal

Investigating manifold-based latent representations for world models, focusing on representation drift and error accumulation during recursive prediction.

May 2026 - present

Research Intern · Lossfunk

Investigating bias localisation in vision models through mathematical and empirical analysis of layers and representation subspaces; developed a regulariser that mitigates multiple attribute biases while preserving predictive performance.

May 2025 - present

Deep Learning Intern · VisDom Lab, IISER Bhopal

Supervised by Dr Vinod K. Kurmi. Researching internal representations of vision and language models, including robustness under distribution shift and multi-step reasoning, and designing methods that address biased representations, failure patterns, fairness, and reliability.

Jul - Aug 2025

Research Intern · TCS Research

Contributed to computer-vision model development and empirical validation; developed adversarial debiasing and representation regularisation for skin-lesion classification across skin-tone groups, leading to the ICVGIP 2025 paper.

2027

Reviewer · ICLR 2027

Serving as a reviewer for the International Conference on Learning Representations.

2026

Reviewer · BMVC 2026

Serving as a reviewer for the British Machine Vision Conference.

2026

Reviewer · ICVGIP 2026

Serving as a reviewer for the Indian Conference on Computer Vision, Graphics and Image Processing.

Background

Education, awards & skills

Department of Data Science and Engineering, Indian Institute of Science Education and Research Bhopal.

Education

  • B.S. in Data ScienceIndian Institute of Science Education and Research Bhopal · Aug 2023 - presentDepartment of Data Science and Engineering · CGPA 9.18/10
  • Pre-University (Physics, Chemistry, Mathematics and Biology)K. H. Patil Pre-University College, Hirekerur · 2020 - 2022Percentage: 90%

Relevant coursework

Machine Learning · Artificial Intelligence · Advanced Deep Learning · Computer Vision · Natural Language Processing · Optimisation · Linear Algebra · Multivariable Calculus · Probability and Statistics

Awards & fellowships

  • SPAR AI Safety Fellowship, 2026Selected for the SPAR fellowship focused on AI Safety and Alignment research and professional development.
  • CODS 2026 Student Travel GrantAwarded a student travel grant to support participation in CODS 2026.
  • Undergraduate Forum @ IndoML 2026Selected to participate in the Undergraduate Forum at IndoML 2026 based on the submitted application.

Workshops & activities

3D Modeling and Cloud Data Processing Workshop, IISER Bhopal (2024) · Model Solvay Conference, IISER Bhopal (2025)

Technical skills

  • Programming and toolsPython, C, C++, PyTorch, TensorFlow, scikit-learn, NumPy, Git
  • MethodsRepresentation learning, self-supervised learning, CNNs, transformers, foundation models, optimisation, generative modelling
  • Research practiceExperimental design, ablation studies, representation analysis, adversarial training, fairness evaluation, distribution-shift analysis
PyTorchTransformersFairnessRobustnessWorld models

Research log

Now & notes

A short, living record of research milestones, ideas, and progress.

Collaborate

Have a research question worth exploring?

mohammedkaif23@iiserb.ac.in ↗