Publications

Papers and preprints on multimodal misinformation detection, agentic reasoning, and interpretability.

Google Scholar arXiv Home
§ 01 · Preprints

Under review

2026
Under review · COLM 2026 arXiv · 2510.17590

MERIT: A Modular Framework for Multimodal Misinformation Detection with Web-Grounded Reasoning

M. N. Sharear Shopnil, S. Duwal, A. Tyagi, A. M. Proma

An agentic pipeline that orchestrates large vision-language models with web search to build verifiable evidence graphs and traceable rationales before committing to a verdict. The point isn't accuracy — it's that you can read the trace and check the work.

arXiv PDF Cite
§ 02 · Conference

Proceedings

2021 → 2026
Accepted · PP-MisDet @ CVPR 2026 arXiv · 2505.18221

Evidence-Grounded Multimodal Misinformation Detection with Attention-Based GNNs

S. Duwal, M. N. Sharear Shopnil, A. Tyagi, A. M. Proma

We formalise claims and evidence as separate graphs and score image-text consistency with a cross-graph attention GNN. The attention maps turn out to be interpretable; the accuracy (93.05%) beats the strongest LLM baseline by 2.82 points.

arXiv PDF
IEEE UEMCON · 2021 Undergrad thesis work

Demystifying Black-box Learning Models of Rumor Detection from Social Media Posts

F. Tafannum, M. N. Sharear Shopnil, A. Salsabil, N. Ahmed, et al.

Applied explainable-AI techniques (LIME, attention attribution) to interpret black-box rumor detection models trained on social media data. The question that started everything: why did the model say that?

DOI
§ 03 · Writing

Technical writing

Medium
2024

Understanding Causal Model Induction in Neural Networks for Interpretability

Medium
2024

You're Not a Fraud: Beating Imposter Syndrome in the Age of Vibe Coding

Medium