Sayan K Chaki
Initializing Forward Pass
Loading weights...
Postdoctoral Researcher· Inria· Laboratoire Hubert Curien· Ecole Centrale Lyon· Naver Labs Europe

Sayan Chaki

Researcher in Machine Mearning
Fairness in LLMs Multi-Agent Systems Equivariant Generative Models Social Choice Theory Computer Vision Deep Learning
View Research Get in Touch
01 — About
Portrait
S
PhD
2025
Introduction

Postdoctoral researcher working within the ANR DIKE project on fairness in Large Language Models and multi-agent resource allocation. My work stands at the intersection of Game Theory, Social Choice Dynamics and Fairness in Agentic LLMs.

My PhD research pioneered advances in equivariant generative architectures, with developing the state-of-the-art model for Equivariant Object Detection.

Inria École Centrale de Lyon Univ. Jean Monnet Naver Labs Europe ANR DIKE Univ. Lyon 2
02 — Research
Publications &
Contributions
2026
Equivariant Unsupervised Object Detection with Learnable Riesz Transform and Composite Spatial Transformers
CVPR 2026 Findings
Sayan Chaki, Thierry Fournel, Rémi Emonet
Accepted
2026
Fairness as Emergence: Alignment and Interaction in Multi-Agent LLM Systems
COLM 2026 (Submitted)
Sayan Chaki, Julien Veclin, Antoine Gourru
In Review
2025
Decoding Ink in Historical Documents: Anomaly Localisation and Source Separation
Workshop on Information Optics (WIO 2025), Barcelona · Springer Nature
Sayan K Chaki, M Hebert, JM Becker, R Emonet, T Fournel
Published
2024
Historical Printed Ornaments: Dataset and Tasks
ICDAR 2024 · Springer Nature Switzerland
Chaki SK, Baltaci ZS, Vincent E, Emonet R, Vial-Bonacci F, Bahier-Porte C, Aubry M, Fournel T
Published
2023
One-to-Many Pattern Comparison Combining Fully-Connected Autoencoder with Spatial Transformer for Ornament Investigation
Research Square Preprint
Sayan Chaki, Simon Steinlin, Rémi Emonet et al.
Preprint
TBD
Localising Perceptual Anomalies with Riesz Encoding and Gabor Skeleton Masks
In Preparation
Sayan Chaki, Rémi Emonet, Thierry Fournel
Submitted
03 — Experience
Work and Internships
Dec 2025 — Present
Postdoctoral Researcher
Inria · École Centrale Lyon · Univ. Jean Monnet · Naver Labs Europe
  • Investigating Fairness in LLMs in Multi-Agent Resource Allocation (ANR DIKE project)
  • Studying deviations in safe and fair models to investigate backdoor attacks in safely aligned language models
  • Leveraging optimisation policies to analyse multi-agent dynamics under Arrow's Impossibility constraints
Sep 2022 — Nov 2025
PhD Scholar
Inria · Laboratoire Hubert Curien · Institut d'Optique Graduate School · UJM
  • Built unsupervised equivariant object detection models for the ANR "Rey's Ornament Image Investigation" project
  • Leveraged group theory and PAC-Bayesian frameworks to mathematically establish model generalisability
  • Demonstrated SOTA on unsupervised object decomposition and anomaly localisation
  • Collaborated with interdisciplinary teams in digital humanities and heritage preservation
Mar 2022 — Jul 2022
Research Intern in Computer Vision
Laboratoire Hubert Curien · UJM · CNRS UMR 5516
  • Designed Spatial Transformer AutoEncoder for anomaly segmentation on MVTec benchmark data
  • Improved SOTA performance for ornament detection and segmentation
May 2021 — Jul 2021
Software Developer Intern
KLA Tencor India
  • Engineered anomaly segmentation algorithms for 3D fault detection
  • Streamlined cross-platform UI ecosystem using React and Electron JS
May 2021 — Jul 2021
Research Assistant · Analytic Number Theory
Indian Statistical Institute, Kolkata · Advised by Satadal Ganguly
  • Results on irreducible Diophantine equations and their connections to longstanding problems in number theory
05 — Invited Talks
September 2025
Methods for Micro and Mesotypography
Herzog August Bibliothek, Wolfenbüttel, Germany
Invited Speaker
November 2024
Intermediality and Computational Humanities Hackathon
University of Vienna
Team Leader · NeRF-based Byzantine Sigillography
June 2024
Generative Modeling Summer School
TU Eindhoven · DTU · Inria
Poster Presenter
March 2024
Computing the Early Page in Modern Europe
University of Oxford
Invited Speaker & Collaborator
June 2023
Google DeepMind ProbAI Summer School
NTNU Trondheim, Norway
Attendee · Probabilistic AI
2024
Deep Learning Enabled Autonomous Drone Network
Pulsalys Explaura Bootcamp · Lyon
🏆 Coup de Cœur Award
06 — Education
Foundations of
a Mind
PhD
Sep 2022 — Nov 2025
Doctor of Philosophy
Image, Vision & Signal Processing
Inria · Laboratoire Hubert Curien · Institut d'Optique Graduate School · IHRIM · UJM · Univ Lyon
MSc
Oct 2020 — Jul 2022
Master of Science
Computer Science
National Institute of Technology, Tiruchirappalli, India
BSc
Jul 2017 — Jun 2020
Bachelor of Science
Statistics
Ramakrishna Mission Residential College (Narendrapur) · University of Calcutta
07 — Contact
Let's Create
Something Beautiful
Whether you want to discuss fairness in AI, collaborate on research, or simply exchange ideas across the boundary of art and intelligence — I am here.
sayanchakiwork@gmail.com LinkedIn GitHub +33 768 753 462