arXiv subjects

Ramana Kompella

Publications and source records attributed to Ramana Kompella.

At least 37 records · Page 2Linked to original sources

arXiv2025

Uncertainty-Aware, Risk-Adaptive Access Control for Agentic Systems using an LLM-Judged TBAC Model

The proliferation of autonomous AI agents within enterprise environments introduces a critical security challenge: managing access control for emergent, novel tasks for which no predefined policies exist. This paper introduces an advanced security framework that extends the Task-Based Access Control (TBAC) model by using a Large Language Model (LLM) as an autonomous, risk-aware judge. This model makes access control decisions not only based on an agent's intent but also by explicitly considering the inherent \textbf{risk associated with target resources} and the LLM's own \textbf{model uncertainty} in its decision-making process. When an agent proposes a novel task, the LLM judge synthesizes a just-in-time policy while also computing a composite risk score for the task and an uncertainty estimate for its own reasoning. High-risk or high-uncertainty requests trigger more stringent controls, such as requiring human approval. This dual consideration of external risk and internal confidence allows the model to enforce a more robust and adaptive version of the principle of least privilege, paving the way for safer and more trustworthy autonomous systems.

cs.CR↗

arXiv2025

A First-order Generative Bilevel Optimization Framework for Diffusion Models

Diffusion models, which iteratively denoise data samples to synthesize high-quality outputs, have achieved empirical success across domains. However, optimizing these models for downstream tasks often involves nested bilevel structures, such as tuning hyperparameters for fine-tuning tasks or noise schedules in training dynamics, where traditional bilevel methods fail due to the infinite-dimensional probability space and prohibitive sampling costs. We formalize this challenge as a generative bilevel optimization problem and address two key scenarios: (1) fine-tuning pre-trained models via an inference-only lower-level solver paired with a sample-efficient gradient estimator for the upper level, and (2) training diffusion model from scratch with noise schedule optimization by reparameterizing the lower-level problem and designing a computationally tractable gradient estimator. Our first-order bilevel framework overcomes the incompatibility of conventional bilevel methods with diffusion processes, offering theoretical grounding and computational practicality. Experiments demonstrate that our method outperforms existing fine-tuning and hyperparameter search baselines.

cs.LG↗

arXiv2025

40Gbps Tri-type Quantum Random Number Generator

Traditional quantum random number generators can produce only one type of random number, while the optimal distribution of random numbers for different applications is usually distinct. The typical solution to this challenge is either using different quantum phenomena for different types of random number, or converting one distribution of random numbers to another type. However, the former solution requires multiple hardware systems, while the latter one sacrifices a lot of secure bits. Here, we develop a quantum random number generator that can on-demand produce three distribution types of random numbers at over 60 Gbits/s (Gbps) raw bits by measuring the quantum vacuum noise. After randomness extraction, over 42 Gbps secure bit rate is demonstrated for uniform random numbers, and over 14 Gbps secure bit rate for Gaussian random number. Due to the lack of Rayleigh randomness extraction, only denoised Rayleigh raw bits are generated. Switching between different types of random numbers is achieved in electronics, which does not affect the generation rate. The random numbers pass NIST and Dieharder tests, and are available for various applications, which can be continuously accessed via Cisco Quantum Random Number web service.

quant-ph↗

arXiv2025

Network-Aware Scheduling for Remote Gate Execution in Quantum Data Centers

Modular quantum computing provides a scalable approach to overcome the limitations of monolithic quantum architectures by interconnecting multiple Quantum Processing Units (QPUs) through a quantum network. In this work, we explore and evaluate two entanglement scheduling strategies-static and dynamic-and analyze their performance in terms of circuit execution delay and network resource utilization under realistic assumptions and practical limitations such as probabilistic entanglement generation, limited communication qubits, photonic switch reconfiguration delays, and topology-induced contention. We show that dynamic scheduling consistently outperforms static scheduling in scenarios with high entanglement parallelism, especially when network resources are scarce. Furthermore, we investigate the impact of communication qubit coherence time, modeled as a cutoff for holding EPR pairs, and demonstrate that aggressive lookahead strategies can degrade performance when coherence times are short, due to premature entanglement discarding and wasted resources. We also identify congestion-free BSM provisioning by profiling peak BSM usage per switch. Our results provide actionable insights for scheduler design and resource provisioning in realistic quantum data centers, bringing system-level considerations closer to practical quantum computing deployment.

quant-ph↗

arXiv2025

A Generative Caching System for Large Language Models

Caching has the potential to be of significant benefit for accessing large language models (LLMs) due to their high latencies which typically range from a small number of seconds to well over a minute. Furthermore, many LLMs charge money for queries; caching thus has a clear monetary benefit. This paper presents a new caching system for improving user experiences with LLMs. In addition to reducing both latencies and monetary costs for accessing LLMs, our system also provides important features that go beyond the performance benefits typically associated with caches. A key feature we provide is generative caching, wherein multiple cached responses can be synthesized to provide answers to queries which have never been seen before. Our generative caches function as repositories of valuable information which can be mined and analyzed. We also improve upon past semantic caching techniques by tailoring the caching algorithms to optimally balance cost and latency reduction with the quality of responses provided. Performance tests indicate that our caches are considerably faster than GPTcache.

cs.DB↗

arXiv2025

Optimized Quantum Circuit Partitioning Across Multiple Quantum Processors

This paper addresses the challenge of scaling quantum computing by employing distributed quantum algorithms across multiple processors. We propose a novel circuit partitioning method that leverages graph partitioning to optimize both qubit and gate teleportation, minimizing the required Einstein-Podolsky-Rosen (EPR) pairs for executing general quantum circuits. Additionally, we formulate an integer linear program to further reduce entanglement requirements by mapping the logical resources of partitioned circuits to the physical constraints of the quantum network. Finally, we analyze the entanglement cost of implementing the Quantum Fourier Transform (QFT) across multiple QPUs, exploiting the circuit's structure to minimize total entanglement consumption.

quant-ph↗

arXiv2025

Layer-Wise Security Framework and Analysis for the Quantum Internet

With its significant security potential, the quantum internet is poised to revolutionize technologies like cryptography and communications. Although it boasts enhanced security over traditional networks, the quantum internet still encounters unique security challenges essential for safeguarding its Confidentiality, Integrity, and Availability (CIA). This study explores these challenges by analyzing the vulnerabilities and the corresponding mitigation strategies across different layers of the quantum internet, including physical, link, network, and application layers. We assess the severity of potential attacks, evaluate the expected effectiveness of mitigation strategies, and identify vulnerabilities within diverse network configurations, integrating both classical and quantum approaches. Our research highlights the dynamic nature of these security issues and emphasizes the necessity for adaptive security measures. The findings underline the need for ongoing research into the security dimension of the quantum internet to ensure its robustness, encourage its adoption, and maximize its impact on society.

cs.NI↗

arXiv2025

Quantum Data Center Infrastructures: A Scalable Architectural Design Perspective

This paper presents the design of scalable quantum networks that utilize optical switches to interconnect multiple quantum processors, facilitating large-scale quantum computing. By leveraging these novel architectures, we aim to address the limitations of current quantum processors and explore the potential of quantum data centers. We provide an in-depth analysis of these architectures through the development of simulation tools and performance metrics, offering a detailed comparison of their advantages and trade-offs. We hope this work serves as a foundation for the development of efficient and resilient quantum networks, designed to meet the evolving demands of future quantum computing applications.

quant-ph↗

arXiv2024

Prompt Diffusion Robustifies Any-Modality Prompt Learning

Foundation models enable prompt-based classifiers for zero-shot and few-shot learning. Nonetheless, the conventional method of employing fixed prompts suffers from distributional shifts that negatively impact generalizability to unseen samples. This paper introduces prompt diffusion, which uses a diffusion model to gradually refine the prompts to obtain a customized prompt for each sample. Specifically, we first optimize a collection of prompts to obtain over-fitted prompts per sample. Then, we propose a prompt diffusion model within the prompt space, enabling the training of a generative transition process from a random prompt to its overfitted prompt. As we cannot access the label of a test image during inference, our model gradually generates customized prompts solely from random prompts using our trained, prompt diffusion. Our prompt diffusion is generic, flexible, and modality-agnostic, making it a simple plug-and-play module seamlessly embedded into existing prompt learning methods for textual, visual, or multi-modal prompt learning. Our diffusion model uses a fast ODE-based sampling strategy to optimize test sample prompts in just five steps, offering a good trade-off between performance improvement and computational efficiency. For all prompt learning methods tested, adding prompt diffusion yields more robust results for base-to-new generalization, cross-dataset generalization, and domain generalization in classification tasks tested over 15 diverse datasets.

cs.LG↗

arXiv2024

Large Language Models Can Learn Temporal Reasoning

While large language models (LLMs) have demonstrated remarkable reasoning capabilities, they are not without their flaws and inaccuracies. Recent studies have introduced various methods to mitigate these limitations. Temporal reasoning (TR), in particular, presents a significant challenge for LLMs due to its reliance on diverse temporal concepts and intricate temporal logic. In this paper, we propose TG-LLM, a novel framework towards language-based TR. Instead of reasoning over the original context, we adopt a latent representation, temporal graph (TG) that enhances the learning of TR. A synthetic dataset (TGQA), which is fully controllable and requires minimal supervision, is constructed for fine-tuning LLMs on this text-to-TG translation task. We confirmed in experiments that the capability of TG translation learned on our dataset can be transferred to other TR tasks and benchmarks. On top of that, we teach LLM to perform deliberate reasoning over the TGs via Chain-of-Thought (CoT) bootstrapping and graph data augmentation. We observed that those strategies, which maintain a balance between usefulness and diversity, bring more reliable CoTs and final results than the vanilla CoT distillation.

cs.CL↗

arXiv2024

Resource Placement for Rate and Fidelity Maximization in Quantum Networks

Existing classical optical network infrastructure cannot be immediately used for quantum network applications due to photon loss. The first step towards enabling quantum networks is the integration of quantum repeaters into optical networks. However, the expenses and intrinsic noise inherent in quantum hardware underscore the need for an efficient deployment strategy that optimizes the allocation of quantum repeaters and memories. In this paper, we present a comprehensive framework for network planning, aiming to efficiently distributing quantum repeaters across existing infrastructure, with the objective of maximizing quantum network utility within an entanglement distribution network. We apply our framework to several cases including a preliminary illustration of a dumbbell network topology and real-world cases of the SURFnet and ESnet. We explore the effect of quantum memory multiplexing within quantum repeaters, as well as the influence of memory coherence time on quantum network utility. We further examine the effects of different fairness assumptions on network planning, uncovering their impacts on real-time network performance.

quant-ph↗

arXiv2024

Multi-tree Quantum Routing in Realistic Topologies

In entanglement distribution networks, communication between two nodes necessitates the generation of end-to-end entanglement by entanglement swapping at intermediate nodes. Efficiently creating end-to-end entanglements over long distances is a key objective. In our prior study on asynchronous routing, we enhanced these entanglement rates by leveraging solely the local knowledge of the entanglement links of a node. This was achieved by creating a tree structure, particularly a destination-oriented directed acyclic graph (DODAG) or a spanning tree, eliminating synchronous operations and conserving unused entanglement links. In this article, we present a multi-tree approach with multiple DODAGs designed to improve end-to-end entanglement rates in large-scale networks, specifically catering to a range of network topologies, including grids and barbells, as well as realistic topologies found in research testbeds like ESnet and Internet2. Our simulations show a marked improvement in end-to-end entanglement rates for specific topologies compared to the single-tree method. This study underscores the promise of asynchronous routing schemes in quantum networks, highlighting the effectiveness of asynchronous routing across different network topologies and proposing a superior routing tactic.

cs.NI↗

arXiv2024

Quantum Key Distribution Networks -- Key Management: A Survey

Secure communication makes the widespread use of telecommunication networks and services possible. With the constant progress of computing and mathematics, new cryptographic methods are being diligently developed. Quantum Key Distribution (QKD) is a promising technology that provides an Information-Theoretically Secure (ITS) solution to the secret-key agreement problem between two remote parties. QKD networks based on trusted repeaters are built to provide service to a larger number of parties at arbitrary distances. They function as an add-on technology to traditional networks, generating, managing, distributing, and supplying ITS cryptographic keys. Since key resources are limited, integrating QKD network services into critical infrastructures necessitates effective key management. As a result, this paper provides a comprehensive review of QKD network key management approaches. They are analyzed to facilitate the identification of potential strategies and accelerate the future development of QKD networks.

cs.CR↗

arXiv2024

Training-Free Semantic Segmentation via LLM-Supervision

Recent advancements in open vocabulary models, like CLIP, have notably advanced zero-shot classification and segmentation by utilizing natural language for class-specific embeddings. However, most research has focused on improving model accuracy through prompt engineering, prompt learning, or fine-tuning with limited labeled data, thereby overlooking the importance of refining the class descriptors. This paper introduces a new approach to text-supervised semantic segmentation using supervision by a large language model (LLM) that does not require extra training. Our method starts from an LLM, like GPT-3, to generate a detailed set of subclasses for more accurate class representation. We then employ an advanced text-supervised semantic segmentation model to apply the generated subclasses as target labels, resulting in diverse segmentation results tailored to each subclass's unique characteristics. Additionally, we propose an assembly that merges the segmentation maps from the various subclass descriptors to ensure a more comprehensive representation of the different aspects in the test images. Through comprehensive experiments on three standard benchmarks, our method outperforms traditional text-supervised semantic segmentation methods by a marked margin.

cs.CV↗

arXiv2023

Adaptive Deep Neural Network Inference Optimization with EENet

Well-trained deep neural networks (DNNs) treat all test samples equally during prediction. Adaptive DNN inference with early exiting leverages the observation that some test examples can be easier to predict than others. This paper presents EENet, a novel early-exiting scheduling framework for multi-exit DNN models. Instead of having every sample go through all DNN layers during prediction, EENet learns an early exit scheduler, which can intelligently terminate the inference earlier for certain predictions, which the model has high confidence of early exit. As opposed to previous early-exiting solutions with heuristics-based methods, our EENet framework optimizes an early-exiting policy to maximize model accuracy while satisfying the given per-sample average inference budget. Extensive experiments are conducted on four computer vision datasets (CIFAR-10, CIFAR-100, ImageNet, Cityscapes) and two NLP datasets (SST-2, AgNews). The results demonstrate that the adaptive inference by EENet can outperform the representative existing early exit techniques. We also perform a detailed visualization analysis of the comparison results to interpret the benefits of EENet.

cs.LG↗

arXiv2023

Causal-DFQ: Causality Guided Data-free Network Quantization

Model quantization, which aims to compress deep neural networks and accelerate inference speed, has greatly facilitated the development of cumbersome models on mobile and edge devices. There is a common assumption in quantization methods from prior works that training data is available. In practice, however, this assumption cannot always be fulfilled due to reasons of privacy and security, rendering these methods inapplicable in real-life situations. Thus, data-free network quantization has recently received significant attention in neural network compression. Causal reasoning provides an intuitive way to model causal relationships to eliminate data-driven correlations, making causality an essential component of analyzing data-free problems. However, causal formulations of data-free quantization are inadequate in the literature. To bridge this gap, we construct a causal graph to model the data generation and discrepancy reduction between the pre-trained and quantized models. Inspired by the causal understanding, we propose the Causality-guided Data-free Network Quantization method, Causal-DFQ, to eliminate the reliance on data via approaching an equilibrium of causality-driven intervened distributions. Specifically, we design a content-style-decoupled generator, synthesizing images conditioned on the relevant and irrelevant factors; then we propose a discrepancy reduction loss to align the intervened distributions of the pre-trained and quantized models. It is worth noting that our work is the first attempt towards introducing causality to data-free quantization problem. Extensive experiments demonstrate the efficacy of Causal-DFQ. The code is available at https://github.com/42Shawn/Causal-DFQ.

cs.CV↗

arXiv2023

Fast and Resource-Efficient Object Tracking on Edge Devices: A Measurement Study

Object tracking is an important functionality of edge video analytic systems and services. Multi-object tracking (MOT) detects the moving objects and tracks their locations frame by frame as real scenes are being captured into a video. However, it is well known that real time object tracking on the edge poses critical technical challenges, especially with edge devices of heterogeneous computing resources. This paper examines the performance issues and edge-specific optimization opportunities for object tracking. We will show that even the well trained and optimized MOT model may still suffer from random frame dropping problems when edge devices have insufficient computation resources. We present several edge specific performance optimization strategies, collectively coined as EMO, to speed up the real time object tracking, ranging from window-based optimization to similarity based optimization. Extensive experiments on popular MOT benchmarks demonstrate that our EMO approach is competitive with respect to the representative methods for on-device object tracking techniques in terms of run-time performance and tracking accuracy. EMO is released on Github at https://github.com/git-disl/EMO.

cs.CV↗

arXiv2023

Mitigating Group Bias in Federated Learning: Beyond Local Fairness

The issue of group fairness in machine learning models, where certain sub-populations or groups are favored over others, has been recognized for some time. While many mitigation strategies have been proposed in centralized learning, many of these methods are not directly applicable in federated learning, where data is privately stored on multiple clients. To address this, many proposals try to mitigate bias at the level of clients before aggregation, which we call locally fair training. However, the effectiveness of these approaches is not well understood. In this work, we investigate the theoretical foundation of locally fair training by studying the relationship between global model fairness and local model fairness. Additionally, we prove that for a broad class of fairness metrics, the global model's fairness can be obtained using only summary statistics from local clients. Based on that, we propose a globally fair training algorithm that directly minimizes the penalized empirical loss. Real-data experiments demonstrate the promising performance of our proposed approach for enhancing fairness while retaining high accuracy compared to locally fair training methods.

cs.LG↗