Sangeetha Abdu Jyothi

Sagar Patel, Junyang Zhang, Sangeetha Abdu Jyothi et al.

Deep Reinforcement Learning (DRL) has shown promise in various networking environments. However, these environments present several fundamental challenges for standard DRL techniques. They are difficult to explore and exhibit high levels of noise and uncertainty. Although these challenges complicate the training process, we find that in practice we can substantially mitigate their effects and even achieve state-of-the-art real-world performance by addressing a factor that has been previously overlooked: the skewed input trace distribution in DRL training datasets. We introduce a generalized framework, Plume, to automatically identify and balance the skew using a three-stage process. First, we identify the critical features that determine the behavior of the traces. Second, we classify the traces into clusters. Finally, we prioritize the salient clusters to improve the overall performance of the controller. Plume seamlessly works across DRL algorithms, without requiring any changes to the DRL workflow. We evaluated Plume on three networking environments, including Adaptive Bitrate Streaming, Congestion Control, and Load Balancing. Plume offers superior performance in both simulation and real-world settings, across different controllers and DRL algorithms. For example, our novel ABR controller, Gelato trained with Plume consistently outperforms prior state-of-the-art controllers on the live streaming platform Puffer for over a year. It is the first controller on the platform to deliver statistically significant improvements in both video quality and stalling, decreasing stalls by as much as 75%.

Sagar Patel, Sangeetha Abdu Jyothi, Nina Narodytska

We present CrystalBox, a novel, model-agnostic, posthoc explainability framework for Deep Reinforcement Learning (DRL) controllers in the large family of input-driven environments which includes computer systems. We combine the natural decomposability of reward functions in input-driven environments with the explanatory power of decomposed returns. We propose an efficient algorithm to generate future-based explanations across both discrete and continuous control environments. Using applications such as adaptive bitrate streaming and congestion control, we demonstrate CrystalBox's capability to generate high-fidelity explanations. We further illustrate its higher utility across three practical use cases: contrastive explanations, network observability, and guided reward design, as opposed to prior explainability techniques that identify salient features.

Alagappan Ramanathan, Eunju Kang, Dongsu Han et al.

Internet measurement faces twin challenges: complex analyses require expert-level orchestration of tools, yet even syntactically correct implementations can have methodological flaws and can be difficult to verify. Democratizing measurement capabilities thus demands automating both workflow generation and verification against methodological standards established through decades of research. We present Airavat, the first agentic framework for Internet measurement workflow generation with systematic verification and validation. Airavat coordinates a set of agents mirroring expert reasoning: three agents handle problem decomposition, solution design, and code implementation, with assistance from a registry of existing tools. Two specialized engines ensure methodological correctness: a Verification Engine evaluates workflows against a knowledge graph encoding five decades of measurement research, while a Validation Engine identifies appropriate validation techniques grounded in established methodologies. Through four Internet measurement case studies, we demonstrate that Airavat (i) generates workflows matching expert-level solutions, (ii) makes sound architectural decisions, (iii) addresses novel problems without ground truth, and (iv) identifies methodological flaws missed by standard execution-based testing.

Nazmul Takbir, Hamidreza Alikhani, Nikil Dutt et al.

Large Language Model (LLM) serving is increasingly constrained by the growing size of the key-value (KV) cache, which scales with both context length and generation length. Prior work shows that attention is dominated by a small subset of critical tokens, yet existing systems struggle to exploit this efficiently without degrading accuracy, especially in long generation. We make a key observation: the temporal stability of these critical tokens varies significantly across KV heads: some heads consistently focus on the same tokens, while others shift frequently. Building on this insight, we introduce FlexiCache, a hierarchical KV-cache management system that leverages the temporal stability of KV heads to reduce GPU memory usage and computation overhead, while preserving model accuracy. FlexiCache classifies KV heads as stable or unstable: it retains all KV-cache pages from unstable heads in GPU memory, whereas for stable heads, it keeps only the top-K pages on the GPU and offloads the rest to host memory. By exploiting temporal stability, FlexiCache performs periodic reranking for stable heads to fetch newly promoted top pages. Implemented atop vLLM, FlexiCache reduces GPU memory footprint for long-context requests by up to 70%, improves offline serving throughput by 1.38-1.55x, and lowers online token latency by 1.6-2.1x, all while maintaining accuracy in long-context, long-generation scenarios.

Alagappan Ramanathan, Eunju Kang, Dongsu Han et al.

Internet measurement research faces an accessibility crisis: complex analyses require custom integration of multiple specialized tools that demands specialized domain expertise. When network disruptions occur, operators need rapid diagnostic workflows spanning infrastructure mapping, routing analysis, and dependency modeling. However, developing these workflows requires specialized knowledge and significant manual effort. We present ArachNet, the first system demonstrating that LLM agents can independently generate measurement workflows that mimics expert reasoning. Our core insight is that measurement expertise follows predictable compositional patterns that can be systematically automated. ArachNet operates through four specialized agents that mirror expert workflow, from problem decomposition to solution implementation. We validate ArachNet with progressively challenging Internet resilience scenarios. The system independently generates workflows that match expert-level reasoning and produce analytical outputs similar to specialist solutions. Generated workflows handle complex multi-framework integration that traditionally requires days of manual coordination. ArachNet lowers barriers to measurement workflow composition by automating the systematic reasoning process that experts use, enabling broader access to sophisticated measurement capabilities while maintaining the technical rigor required for research-quality analysis.

Shaun Christopher Lee, Sangeetha Abdu Jyothi

Backpropagation is inherently sequential across depth, creating an $O(K)$-deep dependency chain that bottlenecks parallel training. While parallel-scan formulations theoretically reduce this depth to $O(\log K)$, they are computationally prohibitive for modern architectures due to the $O(d^3)$ cost of composing full-rank $d\times d$ Jacobians over the entire hidden state. We introduce Latent Bounded Interfaces (LBI), an algorithmic formulation that makes scan-based backpropagation tractable by restricting inter-region communication to a low-dimensional latent interface, $ m_k \in \mathbb{R}^{r}$, where $r \ll d$. This reduces the adjoint recursion to a suffix scan over $r \times r$ Jacobians, cutting per-combine cost from $O(d^3)$ to $O(r^3)$ while preserving exact gradients under the bounded-interface model. We demonstrate that LBI maintains model quality across four architectures (Mamba-2, Mamba-3, Transformer, and a Mamba--Transformer hybrid) at 47--61M block parameters. Interfaces of dimension $r=16$ suffice to preserve training quality within 0.16--0.35 cross entropy of dense baselines. The resulting framework provides an algorithmic foundation for region-parallel training, reducing cross-device backward communication to a single scan over $K$ fixed-size matrices, of approximately 56 KB for our experimental configurations.

Yurun Song, Zhuoyi Yang, Ian G. Harris et al.

Large Language Models (LLMs) are scaling rapidly, creating significant challenges for collaborative server client distributed training, particularly in terms of communication efficiency and computational overheads. To address these challenges, we implement Parameter-efficient Split Learning, which effectively balances efficiency and performance for collaborative training on low-resource devices. To reduce communication overhead in collaborative training, we introduce Adaptive Mixed bit Activation Quantization (AMAQ), a strategy that progressively compresses activations and gradients from high precision (6 to 8 bits) to low precision (3 to 4 bits). AMAQ achieves this by effectively allocating bit budgets across channels based on feature wise and layer wise importance using bit regularization. Under the same bit budgets, AMAQ outperforms fixed-precision approaches, delivering about 2.5% higher generation accuracy and about 1.3% better classification accuracy for models like LLaMA3 8B and Qwen2.5 7B. In addition, it significantly enhances training stability and reducing ultra-low bit representation collapse during the training. Experiments demonstrate that AMAQ integrates effectively into practical multi-machine collaborative training setups, offering superior inference accuracy with only a modest communication overhead for bits adaptation during training. This trade off makes AMAQ a practical and effective solution for collaborative training with minimal communication cost.

Junchen Zhao, Ali Derakhshan, Dushyant Bharadwaj et al.

Large Language Models (LLMs) promise impressive capabilities, yet their multi-billion-parameter scale makes on-device or low-resource deployment prohibitive. Mixed-precision quantization offers a compelling solution, but existing methods struggle when the average precision drops below four bits, as they rely on isolated, layer-specific metrics that overlook critical inter-layer interactions affecting overall performance. In this paper, we propose two innovations to address these limitations. First, we frame the mixed-precision quantization problem as a cooperative game among layers and introduce Shapley-based Progressive Quantization Estimation (SPQE) to efficiently obtain accurate Shapley estimates of layer sensitivities and inter-layer interactions. Second, building upon SPQE, we propose Interaction-aware Mixed-Precision Quantization (IMPQ) which translates these Shapley estimates into a binary quadratic optimization formulation, assigning either 2 or 4-bit precision to layers under strict memory constraints. Comprehensive experiments conducted on Llama-3, Gemma-2, and Qwen-3 models across three independent PTQ backends (Quanto, HQQ, GPTQ) demonstrate IMPQ's scalability and consistently superior performance compared to methods relying solely on isolated metrics. Across average precisions spanning 4 bit down to 2 bit, IMPQ cuts Perplexity by 20 to 80 percent relative to the best baseline, with the margin growing as the bit-width tightens.

Yurun Song, Junchen Zhao, Ian G. Harris et al.

In this paper, we introduce \textbf{Share}d \textbf{Lo}w \textbf{R}ank \textbf{A}daptation (ShareLoRA), a Large Language Model (LLM) fine-tuning technique that balances parameter efficiency, adaptability, and robustness without compromising performance. By strategically sharing the low-rank weight matrices across different layers, ShareLoRA achieves 44\% to 96\% reduction in trainable parameters compared to standard LoRA, alongside a substantial decrease in memory overhead. This efficiency gain scales with model size, making ShareLoRA particularly advantageous for resource-constrained environments. Importantly, ShareLoRA not only maintains model performance but also exhibits robustness in both classification and generation tasks across diverse models, including RoBERTa, GPT-2, and LLaMA series (1, 2, and 3). It consistently outperforms LoRA in zero-shot, few-shot, and continual fine-tuning scenarios, achieving up to 1.2\% average accuracy improvement, and enhanced generalization across domains. In continual learning settings, ShareLoRA achieves 1.2\% higher accuracy on GSM8K, 0.6\% on HumanEval, and 0.5\% on both MMLU and MMLU-Pro. Our results demonstrate that ShareLoRA supports high-quality fine-tuning while offering strong generalization and continual adaptation across various model scales and diverse tasks.

Sayed Hadi Hashemi, Sangeetha Abdu Jyothi, Brighten Godfrey et al.

The method of choice for parameter aggregation in Deep Neural Network (DNN) training, a network-intensive task, is shifting from the Parameter Server model to decentralized aggregation schemes (AllReduce) inspired by theoretical guarantees of better performance. However, current implementations of AllReduce overlook the interdependence of communication and computation, resulting in significant performance degradation. In this paper, we develop Caramel, a system that accelerates decentralized distributed deep learning through model-aware computation scheduling and communication optimizations for AllReduce. Caramel achieves this goal through (a) computation DAG scheduling that expands the feasible window of transfer for each parameter (transfer boundaries), and (b) network optimizations for smoothening of the load including adaptive batching and pipelining of parameter transfers. Caramel maintains the correctness of the dataflow model, is hardware-independent, and does not require any user-level or framework-level changes. We implement Caramel over TensorFlow and show that the iteration time of DNN training can be improved by up to 3.62x in a cloud environment.

Sayed Hadi Hashemi, Sangeetha Abdu Jyothi, Roy H. Campbell

State-of-the-art deep learning systems rely on iterative distributed training to tackle the increasing complexity of models and input data. The iteration time in these communication-heavy systems depends on the computation time, communication time and the extent of overlap of computation and communication. In this work, we identify a shortcoming in systems with graph representation for computation, such as TensorFlow and PyTorch, that result in high variance in iteration time --- random order of received parameters across workers. We develop a system, TicTac, to improve the iteration time by fixing this issue in distributed deep learning with Parameter Servers while guaranteeing near-optimal overlap of communication and computation. TicTac identifies and enforces an order of network transfers which improves the iteration time using prioritization. Our system is implemented over TensorFlow and requires no changes to the model or developer inputs. TicTac improves the throughput by up to $37.7\%$ in inference and $19.2\%$ in training, while also reducing straggler effect by up to $2.3\times$. Our code is publicly available.