Research

Public papers and technical reports.

My current work focuses on multimodal models for document understanding and computer use. This page records public research across multimodal ML, safety, and computer architecture; conference papers and arXiv technical reports are listed together, with preprints explicitly marked. For the continuously updated indexes, see Google Scholar or Semantic Scholar.

2026

  1. Shieldstral
    Shieldstral
    Mistral AI, Gunjan Dhanuka, and  others
    Technical report · arXiv preprint , 2026
    arXiv
  2. Voxtral TTS
    Voxtral TTS
    Mistral AI, Gunjan Dhanuka, and  others
    Technical report · arXiv preprint , 2026
    arXiv
  3. Voxtral Realtime
    Voxtral Realtime
    Mistral AI, Gunjan Dhanuka, and  others
    Technical report · arXiv preprint , 2026
    arXiv

2025

  1. Dynamic Mitigation of Hardware Trojan Induced Black Hole Router Attack in Network-on-Chip
    Dynamic Mitigation of Hardware Trojan Induced Black Hole Router Attack in Network-on-Chip
    Gunjan Dhanuka, Syam Sankar, and John Jose
    In 2025 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2025
    DOI
  2. Distilling Aggregated Knowledge for Weakly-Supervised Video Anomaly Detection
    Distilling Aggregated Knowledge for Weakly-Supervised Video Anomaly Detection
    Jash Dalvi, Ali Dabouei, Gunjan Dhanuka, and 1 more author
    In 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2025
    arXiv DOI