Introducing Paris

World's First Decentralized Trained Open-Weight Diffusion Model

Oct 7 • 

Gin Jiang

, 

Raihan Seraj

, 

Marcos Villagra

, and 

Bidhan Roy

33

2

July 2025

Tiny Tool Use

Open source library for teaching LLMs tool-use

Jul 1 • 

Raihan Seraj

, 

Marcos Villagra

, and 

Bidhan Roy

20

May 2025

Return on Experience (RoE)

A single benchmark that predicts Reinforcement Learning's future.

May 1 • 

Bidhan Roy

18

3

January 2025

ZKLoRA

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

Jan 21 • 

Bidhan Roy

, 

Peter Potash

, and 

Marcos Villagra

27

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

38

September 2024

Machine Unlearning

Right to be forgotten by AI

Sep 10, 2024 • 

Bidhan Roy

 and 

Marcos Villagra

12

June 2024

With Great Data, Comes Great Responsibility II

Privacy preserving machine learning (PPML) with TEE and MPC.

Jun 10, 2024 • 

Bidhan Roy

 and 

Marcos Villagra

12

2

May 2024

Data Synthesis

Synthetic Data 101

May 15, 2024 • 

Marcos Villagra

 and 

Bidhan Roy

3

April 2024

The Inference Interference

How to design verifiable inference that’s not slow

Apr 11, 2024 • 

Bidhan Roy

 and 

Marcos Villagra

6

2

March 2024

With Great Data, Comes Great Responsibility

Privacy preserving machine learning (PPML) at Bagel 🥯

Mar 5, 2024 • 

Bidhan Roy

 and 

Marcos Villagra

8

3