Publications by Rachid Guerraoui

ARGO: Overcoming hardware dependence in distributed learning

Mobile devices offer a valuable resource for distributed learning alongside traditional computers, encouraging energy efficiency and privacy through local computations. However, the hardware limitations of these devices makes it impossible to use classical ...

2025

Partial Synchrony for Free: New Upper Bounds for Byzantine Agreement

Rachid Guerraoui, Manuel José Ribeiro Vidigueira, Jovan Komatovic, Pierre Philippe Civit

Byzantine agreement allows n processes to decide on a common value, in spite of arbitrary failures. The seminal Dolev-Reischuk bound states that any deterministic solution to Byzantine agreement exchanges Ω(n2) bits. In synchronous networks, with a known u ...

Society for Industrial and Applied Mathematics2025

The Utility and Complexity of In- and Out-of-Distribution Machine Unlearning

Rachid Guerraoui, Youssef Allouah

Machine unlearning, the process of selectively removing data from trained models, is increasingly crucial for addressing privacy concerns and knowledge gaps post-deployment. Despite this importance, existing approaches are often heuristic and lack formal g ...

ICLR2025

Repeated Agreement is Cheap! On Weak Accountability and Multishot Byzantine Agreement

Rachid Guerraoui, Manuel José Ribeiro Vidigueira, Jovan Komatovic, Pierre Philippe Civit

Byzantine Agreement (BA) allows 𝑛 processes to propose input values to reach consensus on a common, valid 𝐿 𝑜-bit value, even in the presence of up to 𝑡 < 𝑛 faulty processes that can deviate arbitrarily from the protocol. Although strategies like randomiza ...

ACM2025

Efficient Federated Search for Retrieval-Augmented Generation

Rachid Guerraoui, Anne-Marie Kermarrec, Diana Andreea Petrescu, Rafael Pereira Pires, Marinus Abraham de Vos

Large language models (LLMs) have demonstrated remarkable capabilities across various domains but remain susceptible to hallucinations and inconsistencies, limiting their reliability. Retrieval-augmented generation (RAG) mitigates these issues by grounding ...

ACM2025

Evaluating Blockchain Fault Tolerance with Stabl

Rachid Guerraoui, Gauthier Jérôme Timothée Voron

Despite their distributed nature, blockchains frequently experience failures that disrupt availability, delay transactions, and, in some cases, cause total network outages. Ensuring fault tolerance is critical for the reliability of blockchain-based applic ...

IEEE2025

Efficient Signature-Free Validated Agreement

Rachid Guerraoui, Manuel José Ribeiro Vidigueira, Jovan Komatovic, Pierre Philippe Civit

Byzantine agreement enables n processes to agree on a common L-bit value, despite up to t > 0 arbitrary failures. A long line of work has been dedicated to improving the bit complexity of Byzantine agreement in synchrony. This has culminated in COOL, an er ...

Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing2024

Robust Sparse Voting

Rachid Guerraoui, Le Nguyen Hoang, Youssef Allouah, Oscar Jean Olivier Villemaud

Many applications, such as content moderation and recommendation, require reviewing and scoring a large number of alternatives. Doing so robustly is however very challenging. Indeed, voters' inputs are inevitably sparse: most alternatives are only scored b ...

2024

Rapidly verifiable aggregate signatures

Rachid Guerraoui, Matteo Monti, Manuel José Ribeiro Vidigueira, Martina Camaioni, Gauthier Jérôme Timothée Voron, Pierre-Louis Blaise Roman

A method for aggregating digital signatures comprises the following steps carried out by a signature aggregator: receiving first data packages from signers, each first data package comprising a signer identifier, a payload, and a payload signature; verifyi ...

2024

Robust Machine Learning - Distributed Methods for Safe AI

Rachid Guerraoui, Nirupam Gupta, Rafaël Benjamin Pinot

Today, machine learning algorithms are often distributed across multiple machines to leverage more computing power and more data. However, the use of a distributed framework entails a variety of security threats. In particular, some of the machines may mis ...

Springer2024

Robust and Private Federated Learning on LLMs

Rachid Guerraoui, Nirupam Gupta

Large Language Models (LLMs) have gained significant attention in recent years due to their potential to revolutionize various industries and sectors. However, scaling LLMs further requires access to substantial linguistic resources that are being rapidly ...

Springer Nature2024

Revisiting Ensembling in One-Shot Federated Learning

Rachid Guerraoui, Anne-Marie Kermarrec, Akash Balasaheb Dhasade, Rafael Pereira Pires, Nirupam Gupta, Rafaël Benjamin Pinot, Rishi Sharma, Youssef Allouah

Federated learning (FL) is an appealing approach to training machine learning models without sharing raw data. However, standard FL algorithms are iterative and thus induce a significant communication cost. One-shot federated learning (OFL) trades the iter ...

Curran Associates, Inc.2024

Brief Announcement: A Case for Byzantine Machine Learning

Rachid Guerraoui, Sadegh Farhadkhani, Nirupam Gupta, Rafaël Benjamin Pinot

The success of machine learning (ML) has been intimately linked with the availability of large amounts of data, typically collected from heterogeneous sources and processed on vast networks of computing devices (also called workers). Beyond accuracy, the u ...

ACM2024

Accelerating Transfer Learning with Near-Data Computation on Cloud Object Stores

Rachid Guerraoui, Diana Andreea Petrescu, Arsany Hany Abdelmessih Guirguis

Storage disaggregation underlies today's cloud and is naturally complemented by pushing down some computation to storage, thus mitigating the potential network bottleneck between the storage and compute tiers. We show how ML training benefits from storage ...

Association for Computing Machinery, Inc2024

All Byzantine Agreement Problems Are Expensive

Rachid Guerraoui, Seth Gilbert, Manuel José Ribeiro Vidigueira, Jovan Komatovic, Pierre Philippe Civit

Byzantine agreement, arguably the most fundamental problem in distributed computing, operates among n processes, out of which t < n can exhibit arbitrary failures. The problem states that all correct (non-faulty) processes must eventually decide (terminati ...

ACM2024

Adversarial Evasion on LLMs

Rachid Guerraoui

While Machine Learning (ML) applications have shown impressive achievements in tasks such as computer vision, NLP, and control problems, such achievements were possible, first and foremost, in the best-case-scenario setting. Unfortunately, settings where M ...

Springer Nature2024

SWARM: Replicating Shared Disaggregated-Memory Data in No Time

Rachid Guerraoui, Mihail Igor Zablotchi, Athanasios Xygkis, Antoine Murat

Memory disaggregation is an emerging data center architecture that improves resource utilization and scalability. Replication is key to ensure the fault tolerance of applications, but replicating shared data in disaggregated memory is hard. We propose SWAR ...

ACM2024

Chop Chop: Byzantine Atomic Broadcast to the Network Limit

Rachid Guerraoui, Matteo Monti, Manuel José Ribeiro Vidigueira, Martina Camaioni, Gauthier Jérôme Timothée Voron

At the heart of state machine replication, the celebrated technique enabling decentralized and secure universal computation, lies Atomic Broadcast, a fundamental communication primitive that orders, authenticates, and deduplicates messages. This paper pres ...

USENIX2024

DARE to Agree: Byzantine Agreement With Optimal Resilience and Adaptive Communication

Rachid Guerraoui, Seth Gilbert, Manuel José Ribeiro Vidigueira, Jovan Komatovic, Pierre Philippe Civit

Byzantine Agreement (BA) enables n processes to reach consensus on a common valid Lo-bit value, even in the presence of up to t < n faulty processes that can deviate arbitrarily from their prescribed protocol. Despite its significance, the optimal communic ...

ACM2024

Towards Practical Homomorphic Aggregation in Byzantine-Resilient Distributed Learning

Rachid Guerraoui, John Stephan

The growing availability of distributed data has led to the increased use of machine learning (ML) algorithms in distributed topologies, where multiple nodes collaborate to train models under the coordination of a central server. However, distributed learn ...

Association for Computing Machinery, Inc2024