Introducing Paris 2.0

A Decentralized Diffusion Model for Video Generation

May 28 • Ali Rouzbayani, Bidhan Roy, Marcos Villagra, and Gin Jiang

13

1

March 2026

Heterogeneous Decentralized Diffusion Models

Training free merge of heterogeneous objective diffusion ensemble

Mar 11 • Gin Jiang and Bidhan Roy

23

1

2

February 2026

Stability–Quality Paradox in Decentralized Diffusion Models

Why the most stable samplers in decentralized diffusion produce the worst generations

Feb 6 • Marcos Villagra, Bidhan Roy, Raihan Seraj, and Gin Jiang

22

4

3

October 2025

Introducing Paris

World's First Decentralized Trained Open-Weight Diffusion Model

Oct 7, 2025 • Gin Jiang, Raihan Seraj, Marcos Villagra, and Bidhan Roy

40

3

3

July 2025

Tiny Tool Use

Open source library for teaching LLMs tool-use

Jul 1, 2025 • Raihan Seraj, Marcos Villagra, and Bidhan Roy

21

1

May 2025

Return on Experience (RoE)

A single benchmark that predicts Reinforcement Learning's future.

May 1, 2025 • Bidhan Roy

19

3

3

January 2025

ZKLoRA

Zero-Knowledge Verification of LoRA training on SOTA Models in 1-2 Seconds

Jan 21, 2025 • Bidhan Roy, Peter Potash, and Marcos Villagra

29

4

November 2024

Train Fast, But Think Slow

Inference Time Compute vs Training Time Compute for AI Reasoning

Nov 6, 2024 • Bidhan Roy and Marcos Villagra

39

1