arXiv subjects

Matthew D. Laws

Publications and source records attributed to Matthew D. Laws.

3 recordsLinked to original sources

arXiv2026

APWA: A Distributed Architecture for Parallelizable Agentic Workflows

Autonomous multi-agent systems based on large language models (LLMs) have demonstrated remarkable abilities in independently solving complex tasks in a wide breadth of application domains. However, these systems hit critical reasoning, coordination, and computational scaling bottlenecks as the size and complexity of their tasks grow. These limitations hinder multi-agent systems from achieving high-throughput processing for highly parallelizable tasks, despite the availability of parallel computing and reasoning primitives in the underlying LLMs. We introduce the Agent-Parallel Workload Architecture (APWA), a distributed multi-agent system architecture designed for the efficient processing of heavily parallelizable agentic workloads. APWA facilitates parallel execution by decomposing workflows into non-interfering subproblems that can be processed using independent resources without cross-communication. It supports heterogeneous data and parallel processing patterns, and it accommodates tasks from a wide breadth of domains. In our evaluation, we demonstrate that APWA can dynamically decompose complex queries into parallelizable workflows and scales on larger tasks in settings where prior systems fail completely.

cs.AI↗

arXiv2026

Attacks and Mitigations for Distributed Governance of Agentic AI under Byzantine Adversaries

Agentic AI governance is a critical component of agentic AI infrastructure ensuring that agents follow their owner's communication and interaction policies, and providing protection against attacks from malicious agents. The state-of-the-art solution, SAGA, assumes a logically centralized point of trust, the Provider, which serves as a repository for user and agent information and actively enforces policies. While SAGA provides protection against malicious agents, it remains vulnerable to a malicious Provider that deviates from the protocol, undermining the security of the identity and access control infrastructure. Deployment on both private and public clouds, each susceptible to insider threats, further increases the risk of Provider compromise. In this work, we analyze the attacks that can be mounted from a compromised Provider, taking into account the different system components and realistic deployments. We identify and execute several concrete attacks with devastating effects: undermining agent attributability, extracting private data, or bypassing access control. We then present three types of solutions for securing the Provider that offer different trade-offs between security and performance. We first present SAGA-BFT, a fully byzantine-resilient architecture that provides the strongest protection, but incurs significant performance degradation, due to the high-cost of byzantine resilient protocols. We then propose SAGA-MON and SAGA-AUD, two novel solutions that leverage lightweight server-side monitoring or client-side auditing to provide protection against most classes of attacks with minimal overhead. Finally, we propose SAGA-HYB, a hybrid architecture that combines byzantine-resilience with monitoring and auditing to trade-off security for performance. We evaluate all the architectures and compare them with SAGA. We discuss which solution is best and under what conditions.

cs.CR↗

arXiv2024

SPIDER: Improved Succinct Rank and Select Performance

Rank and select data structures seek to preprocess a bit vector to quickly answer two kinds of queries: rank(i) gives the number of 1 bits in slots 0 through i, and select(j) gives the first slot s with rank(s) = j. A succinct data structure can answer these queries while using space much smaller than the size of the original bit vector. State of the art succinct rank and select data structures use as little as 4% extra space while answering rank and select queries quickly. Rank queries can be answered using only a handful of array accesses. Select queries can be answered by starting with similar array accesses, followed by a linear scan. Despite these strong results, a tradeoff remains: data structures that use under 4% space are significantly slower at answering rank and select queries than less-space-efficient data structures (using, say, > 20% extra space). In this paper we make significant progress towards closing this gap. We give a new data structure, SPIDER, which uses 3.82% extra space. SPIDER gives the best rank query time for data sets of 8 billion or more bits, even compared to less space-efficient data structures. For select queries, SPIDER outperforms all data structures that use less than 4% space, and significantly closes the gap in select performance between data structures with less than 4% space, and those that use more (over 20%) space. SPIDER makes two main technical contributions. For rank queries, it improves performance by interleaving the metadata with the bit vector to improve cache efficiency. For select queries, it uses predictions to almost eliminate the cost of the linear scan. These predictions are inspired by recent results on data structures with machine-learned predictions, adapted to the succinct data structure setting. Our results hold on both real and synthetic data, showing that these predictions are effective in practice.

cs.DS↗