arXiv subjects

Sairam Sundaresan

Publications and source records attributed to Sairam Sundaresan.

At least 19 recordsLinked to original sources

arXiv2025

HiSAXy: A fast methodology for solar wind structure identification in millions of time series

We present a hybridized unsupervised clustering algorithm Hisaxy as a novel way to identify frequently occurring magnetic structures embedded in the interplanetary magnetic field (IMF) carried by the solar wind. The Hisaxy algorithm utilizes a combination of indexable Symbolic Aggregate approXimation (iSAX) and Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) to efficiently identify clusters of patterns embedded in time series data. We utilized Hisaxy to identify small-scale structures, known as discontinuities, embedded in time series measurements of the IMF. In doing so, we demonstrate the capability of the algorithm to significantly reduce the amount of human analysis hours required to identify these structures, all the while maintaining a high degree of self similarity within a given cluster of time series data.

astro-ph.SR↗

arXiv2025

CIPHER: Scalable Time Series Analysis for Physical Sciences with Application to Solar Wind Phenomena

Labeling or classifying time series is a persistent challenge in the physical sciences, where expert annotations are scarce, costly, and often inconsistent. Yet robust labeling is essential to enable machine learning models for understanding, prediction, and forecasting. We present the \textit{Clustering and Indexation Pipeline with Human Evaluation for Recognition} (CIPHER), a framework designed to accelerate large-scale labeling of complex time series in physics. CIPHER integrates \textit{indexable Symbolic Aggregate approXimation} (iSAX) for interpretable compression and indexing, density-based clustering (HDBSCAN) to group recurring phenomena, and a human-in-the-loop step for efficient expert validation. Representative samples are labeled by domain scientists, and these annotations are propagated across clusters to yield systematic, scalable classifications. We evaluate CIPHER on the task of classifying solar wind phenomena in OMNI data, a central challenge in space weather research, showing that the framework recovers meaningful phenomena such as coronal mass ejections and stream interaction regions. Beyond this case study, CIPHER highlights a general strategy for combining symbolic representations, unsupervised learning, and expert knowledge to address label scarcity in time series across the physical sciences. The code and configuration files used in this study are publicly available to support reproducibility.

cs.LG↗

arXiv2024

LLaMA-NAS: Efficient Neural Architecture Search for Large Language Models

The abilities of modern large language models (LLMs) in solving natural language processing, complex reasoning, sentiment analysis and other tasks have been extraordinary which has prompted their extensive adoption. Unfortunately, these abilities come with very high memory and computational costs which precludes the use of LLMs on most hardware platforms. To mitigate this, we propose an effective method of finding Pareto-optimal network architectures based on LLaMA2-7B using one-shot NAS. In particular, we fine-tune LLaMA2-7B only once and then apply genetic algorithm-based search to find smaller, less computationally complex network architectures. We show that, for certain standard benchmark tasks, the pre-trained LLaMA2-7B network is unnecessarily large and complex. More specifically, we demonstrate a 1.5x reduction in model size and 1.3x speedup in throughput for certain tasks with negligible drop in accuracy. In addition to finding smaller, higher-performing network architectures, our method does so more effectively and efficiently than certain pruning or sparsification techniques. Finally, we demonstrate how quantization is complementary to our method and that the size and complexity of the networks we find can be further decreased using quantization. We believe that our work provides a way to automatically create LLMs which can be used on less expensive and more readily available hardware platforms.

cs.AI↗

arXiv2024

SuNeRF: 3D reconstruction of the solar EUV corona using Neural Radiance Fields

To understand its evolution and the effects of its eruptive events, the Sun is permanently monitored by multiple satellite missions. The optically-thin emission of the solar plasma and the limited number of viewpoints make it challenging to reconstruct the geometry and structure of the solar atmosphere; however, this information is the missing link to understand the Sun as it is: a three-dimensional evolving star. We present a method that enables a complete 3D representation of the uppermost solar layer (corona) observed in extreme ultraviolet (EUV) light. We use a deep learning approach for 3D scene representation that accounts for radiative transfer, to map the entire solar atmosphere from three simultaneous observations. We demonstrate that our approach provides unprecedented reconstructions of the solar poles, and directly enables height estimates of coronal structures, solar filaments, coronal hole profiles, and coronal mass ejections. We validate the approach using model-generated synthetic EUV images, finding that our method accurately captures the 3D geometry of the Sun even from a limited number of 32 ecliptic viewpoints ($|\text{latitude}| \leq 7^\circ$). We quantify uncertainties of our model using an ensemble approach that allows us to estimate the model performance in absence of a ground-truth. Our method enables a novel view of our closest star, and is a breakthrough technology for the efficient use of multi-instrument datasets, which paves the way for future cluster missions.

astro-ph.SR↗

arXiv2023

SimQ-NAS: Simultaneous Quantization Policy and Neural Architecture Search

Recent one-shot Neural Architecture Search algorithms rely on training a hardware-agnostic super-network tailored to a specific task and then extracting efficient sub-networks for different hardware platforms. Popular approaches separate the training of super-networks from the search for sub-networks, often employing predictors to alleviate the computational overhead associated with search. Additionally, certain methods also incorporate the quantization policy within the search space. However, while the quantization policy search for convolutional neural networks is well studied, the extension of these methods to transformers and especially foundation models remains under-explored. In this paper, we demonstrate that by using multi-objective search algorithms paired with lightly trained predictors, we can efficiently search for both the sub-network architecture and the corresponding quantization policy and outperform their respective baselines across different performance objectives such as accuracy, model size, and latency. Specifically, we demonstrate that our approach performs well across both uni-modal (ViT and BERT) and multi-modal (BEiT-3) transformer-based architectures as well as convolutional architectures (ResNet). For certain networks, we demonstrate an improvement of up to $4.80x$ and $3.44x$ for latency and model size respectively, without degradation in accuracy compared to the fully quantized INT8 baselines.

cs.LG↗

arXiv2023

Sensi-BERT: Towards Sensitivity Driven Fine-Tuning for Parameter-Efficient BERT

Large pre-trained language models have recently gained significant traction due to their improved performance on various down-stream tasks like text classification and question answering, requiring only few epochs of fine-tuning. However, their large model sizes often prohibit their applications on resource-constrained edge devices. Existing solutions of yielding parameter-efficient BERT models largely rely on compute-exhaustive training and fine-tuning. Moreover, they often rely on additional compute heavy models to mitigate the performance gap. In this paper, we present Sensi-BERT, a sensitivity driven efficient fine-tuning of BERT models that can take an off-the-shelf pre-trained BERT model and yield highly parameter-efficient models for downstream tasks. In particular, we perform sensitivity analysis to rank each individual parameter tensor, that then is used to trim them accordingly during fine-tuning for a given parameter or FLOPs budget. Our experiments show the efficacy of Sensi-BERT across different downstream tasks including MNLI, QQP, QNLI, SST-2 and SQuAD, showing better performance at similar or smaller parameter budget compared to various alternatives.

cs.CL↗

arXiv2023

InstaTune: Instantaneous Neural Architecture Search During Fine-Tuning

One-Shot Neural Architecture Search (NAS) algorithms often rely on training a hardware agnostic super-network for a domain specific task. Optimal sub-networks are then extracted from the trained super-network for different hardware platforms. However, training super-networks from scratch can be extremely time consuming and compute intensive especially for large models that rely on a two-stage training process of pre-training and fine-tuning. State of the art pre-trained models are available for a wide range of tasks, but their large sizes significantly limits their applicability on various hardware platforms. We propose InstaTune, a method that leverages off-the-shelf pre-trained weights for large models and generates a super-network during the fine-tuning stage. InstaTune has multiple benefits. Firstly, since the process happens during fine-tuning, it minimizes the overall time and compute resources required for NAS. Secondly, the sub-networks extracted are optimized for the target task, unlike prior work that optimizes on the pre-training objective. Finally, InstaTune is easy to "plug and play" in existing frameworks. By using multi-objective evolutionary search algorithms along with lightly trained predictors, we find Pareto-optimal sub-networks that outperform their respective baselines across different performance objectives such as accuracy and MACs. Specifically, we demonstrate that our approach performs well across both unimodal (ViT and BERT) and multi-modal (BEiT-3) transformer based architectures.

cs.LG↗

arXiv2022

Sparse Mixture Once-for-all Adversarial Training for Efficient In-Situ Trade-Off Between Accuracy and Robustness of DNNs

Existing deep neural networks (DNNs) that achieve state-of-the-art (SOTA) performance on both clean and adversarially-perturbed images rely on either activation or weight conditioned convolution operations. However, such conditional learning costs additional multiply-accumulate (MAC) or addition operations, increasing inference memory and compute costs. To that end, we present a sparse mixture once for all adversarial training (SMART), that allows a model to train once and then in-situ trade-off between accuracy and robustness, that too at a reduced compute and parameter overhead. In particular, SMART develops two expert paths, for clean and adversarial images, respectively, that are then conditionally trained via respective dedicated sets of binary sparsity masks. Extensive evaluations on multiple image classification datasets across different models show SMART to have up to 2.72x fewer non-zero parameters costing proportional reduction in compute overhead, while yielding SOTA accuracy-robustness trade-off. Additionally, we present insightful observations in designing sparse masks to successfully condition on both clean and perturbed images.

cs.CV↗

arXiv2022

SuNeRF: Validation of a 3D Global Reconstruction of the Solar Corona Using Simulated EUV Images

Extreme Ultraviolet (EUV) light emitted by the Sun impacts satellite operations and communications and affects the habitability of planets. Currently, EUV-observing instruments are constrained to viewing the Sun from its equator (i.e., ecliptic), limiting our ability to forecast EUV emission for other viewpoints (e.g. solar poles), and to generalize our knowledge of the Sun-Earth system to other host stars. In this work, we adapt Neural Radiance Fields (NeRFs) to the physical properties of the Sun and demonstrate that non-ecliptic viewpoints could be reconstructed from observations limited to the solar ecliptic. To validate our approach, we train on simulations of solar EUV emission that provide a ground truth for all viewpoints. Our model accurately reconstructs the simulated 3D structure of the Sun, achieving a peak signal-to-noise ratio of 43.3 dB and a mean absolute relative error of 0.3\% for non-ecliptic viewpoints. Our method provides a consistent 3D reconstruction of the Sun from a limited number of viewpoints, thus highlighting the potential to create a virtual instrument for satellite observations of the Sun. Its extension to real observations will provide the missing link to compare the Sun to other stars and to improve space-weather forecasting.

astro-ph.SR↗

arXiv2022

A Hardware-Aware Framework for Accelerating Neural Architecture Search Across Modalities

Recent advances in Neural Architecture Search (NAS) such as one-shot NAS offer the ability to extract specialized hardware-aware sub-network configurations from a task-specific super-network. While considerable effort has been employed towards improving the first stage, namely, the training of the super-network, the search for derivative high-performing sub-networks is still under-explored. Popular methods decouple the super-network training from the sub-network search and use performance predictors to reduce the computational burden of searching on different hardware platforms. We propose a flexible search framework that automatically and efficiently finds optimal sub-networks that are optimized for different performance metrics and hardware configurations. Specifically, we show how evolutionary algorithms can be paired with lightly trained objective predictors in an iterative cycle to accelerate architecture search in a multi-objective setting for various modalities including machine translation and image classification.

cs.LG↗

arXiv2022

A Fast and Efficient Conditional Learning for Tunable Trade-Off between Accuracy and Robustness

Existing models that achieve state-of-the-art (SOTA) performance on both clean and adversarially-perturbed images rely on convolution operations conditioned with feature-wise linear modulation (FiLM) layers. These layers require many new parameters and are hyperparameter sensitive. They significantly increase training time, memory cost, and potential latency which can prove costly for resource-limited or real-time applications. In this paper, we present a fast learnable once-for-all adversarial training (FLOAT) algorithm, which instead of the existing FiLM-based conditioning, presents a unique weight conditioned learning that requires no additional layer, thereby incurring no significant increase in parameter count, training time, or network latency compared to standard adversarial training. In particular, we add configurable scaled noise to the weight tensors that enables a trade-off between clean and adversarial performance. Extensive experiments show that FLOAT can yield SOTA performance improving both clean and perturbed image classification by up to ~6% and ~10%, respectively. Moreover, real hardware measurement shows that FLOAT can reduce the training time by up to 1.43x with fewer model parameters of up to 1.47x on iso-hyperparameter settings compared to the FiLM-based alternatives. Additionally, to further improve memory efficiency we introduce FLOAT sparse (FLOATS), a form of non-iterative model pruning and provide detailed empirical analysis to provide a three way accuracy-robustness-complexity trade-off for these new class of pruned conditionally trained models.

cs.CV↗

arXiv2022

Deep-SWIM: A few-shot learning approach to classify Solar WInd Magnetic field structures

The solar wind consists of charged particles ejected from the Sun into interplanetary space and towards Earth. Understanding the magnetic field of the solar wind is crucial for predicting future space weather and planetary atmospheric loss. Compared to large-scale magnetic events, smaller-scale structures like magnetic discontinuities are hard to detect but entail important information on the evolution of the solar wind. A lack of labeled data makes an automated detection of these discontinuities challenging. We propose Deep-SWIM, an approach leveraging advances in contrastive learning, pseudo-labeling and online hard example mining to robustly identify discontinuities in solar wind magnetic field data. Through a systematic ablation study, we show that we can accurately classify discontinuities despite learning from only limited labeled data. Additionally, we show that our approach generalizes well and produces results that agree with expert hand-labeling.

astro-ph.SR↗

arXiv2022

A Hardware-Aware System for Accelerating Deep Neural Network Optimization

Recent advances in Neural Architecture Search (NAS) which extract specialized hardware-aware configurations (a.k.a. "sub-networks") from a hardware-agnostic "super-network" have become increasingly popular. While considerable effort has been employed towards improving the first stage, namely, the training of the super-network, the search for derivative high-performing sub-networks is still largely under-explored. For example, some recent network morphism techniques allow a super-network to be trained once and then have hardware-specific networks extracted from it as needed. These methods decouple the super-network training from the sub-network search and thus decrease the computational burden of specializing to different hardware platforms. We propose a comprehensive system that automatically and efficiently finds sub-networks from a pre-trained super-network that are optimized to different performance metrics and hardware configurations. By combining novel search tactics and algorithms with intelligent use of predictors, we significantly decrease the time needed to find optimal sub-networks from a given super-network. Further, our approach does not require the super-network to be refined for the target task a priori, thus allowing it to interface with any super-network. We demonstrate through extensive experiments that our system works seamlessly with existing state-of-the-art super-network training methods in multiple domains. Moreover, we show how novel search tactics paired with evolutionary algorithms can accelerate the search process for ResNet50, MobileNetV3 and Transformer while maintaining objective space Pareto front diversity and demonstrate an 8x faster search result than the state-of-the-art Bayesian optimization WeakNAS approach.

cs.AI↗

arXiv2022

TrimBERT: Tailoring BERT for Trade-offs

Models based on BERT have been extremely successful in solving a variety of natural language processing (NLP) tasks. Unfortunately, many of these large models require a great deal of computational resources and/or time for pre-training and fine-tuning which limits wider adoptability. While self-attention layers have been well-studied, a strong justification for inclusion of the intermediate layers which follow them remains missing in the literature. In this work, we show that reducing the number of intermediate layers in BERT-Base results in minimal fine-tuning accuracy loss of downstream tasks while significantly decreasing model size and training time. We further mitigate two key bottlenecks, by replacing all softmax operations in the self-attention layers with a computationally simpler alternative and removing half of all layernorm operations. This further decreases the training time while maintaining a high level of fine-tuning accuracy.

cs.CL↗

arXiv2020

AttentionLite: Towards Efficient Self-Attention Models for Vision

We propose a novel framework for producing a class of parameter and compute efficient models called AttentionLitesuitable for resource-constrained applications. Prior work has primarily focused on optimizing models either via knowledge distillation or pruning. In addition to fusing these two mechanisms, our joint optimization framework also leverages recent advances in self-attention as a substitute for convolutions. We can simultaneously distill knowledge from a compute-heavy teacher while also pruning the student model in a single pass of training thereby reducing training and fine-tuning times considerably. We evaluate the merits of our proposed approach on the CIFAR-10, CIFAR-100, and Tiny-ImageNet datasets. Not only do our AttentionLite models significantly outperform their unoptimized counterparts in accuracy, we find that in some cases, that they perform almost as well as their compute-heavy teachers while consuming only a fraction of the parameters and FLOPs. Concretely, AttentionLite models can achieve upto30x parameter efficiency and 2x computation efficiency with no significant accuracy drop compared to their teacher.

cs.CV↗

arXiv2020

Attention-based Image Upsampling

Convolutional layers are an integral part of many deep neural network solutions in computer vision. Recent work shows that replacing the standard convolution operation with mechanisms based on self-attention leads to improved performance on image classification and object detection tasks. In this work, we show how attention mechanisms can be used to replace another canonical operation: strided transposed convolution. We term our novel attention-based operation attention-based upsampling since it increases/upsamples the spatial dimensions of the feature maps. Through experiments on single image super-resolution and joint-image upsampling tasks, we show that attention-based upsampling consistently outperforms traditional upsampling methods based on strided transposed convolution or based on adaptive filters while using fewer parameters. We show that the inherent flexibility of the attention mechanism, which allows it to use separate sources for calculating the attention coefficients and the attention targets, makes attention-based upsampling a natural choice when fusing information from multiple image modalities.

cs.CV↗

arXiv2020

RotNet: Fast and Scalable Estimation of Stellar Rotation Periods Using Convolutional Neural Networks

Magnetic activity in stars manifests as dark spots on their surfaces that modulate the brightness observed by telescopes. These light curves contain important information on stellar rotation. However, the accurate estimation of rotation periods is computationally expensive due to scarce ground truth information, noisy data, and large parameter spaces that lead to degenerate solutions. We harness the power of deep learning and successfully apply Convolutional Neural Networks to regress stellar rotation periods from Kepler light curves. Geometry-preserving time-series to image transformations of the light curves serve as inputs to a ResNet-18 based architecture which is trained through transfer learning. The McQuillan catalog of published rotation periods is used as ansatz to groundtruth. We benchmark the performance of our method against a random forest regressor, a 1D CNN, and the Auto-Correlation Function (ACF) - the current standard to estimate rotation periods. Despite limiting our input to fewer data points (1k), our model yields more accurate results and runs 350 times faster than ACF runs on the same number of data points and 10,000 times faster than ACF runs on 65k data points. With only minimal feature engineering our approach has impressive accuracy, motivating the application of deep learning to regress stellar parameters on an even larger scale

astro-ph.SR↗

arXiv2020

Logic2Text: High-Fidelity Natural Language Generation from Logical Forms

Previous works on Natural Language Generation (NLG) from structured data have primarily focused on surface-level descriptions of record sequences. However, for complex structured data, e.g., multi-row tables, it is often desirable for an NLG system to describe interesting facts from logical inferences across records. If only provided with the table, it is hard for existing models to produce controllable and high-fidelity logical generations. In this work, we formulate logical level NLG as generation from logical forms in order to obtain controllable, high-fidelity, and faithful generations. We present a new large-scale dataset, \textsc{Logic2Text}, with 10,753 descriptions involving common logic types paired with the underlying logical forms. The logical forms show diversified graph structure of free schema, which poses great challenges on the model's ability to understand the semantics. We experiment on (1) Fully-supervised training with the full datasets, and (2) Few-shot setting, provided with hundreds of paired examples; We compare several popular generation models and analyze their performances. We hope our dataset can encourage research towards building an advanced NLG system capable of natural, faithful, and human-like generation. The dataset and code are available at https://github.com/czyssrs/Logic2Text.

cs.CL↗