arXiv subjects

Aman Gupta

Publications and source records attributed to Aman Gupta.

At least 37 records · Page 2Linked to original sources

arXiv2025

Obfuscated Quantum and Post-Quantum Cryptography

In this work, we present an experimental deployment of a new design for combined quantum key distribution (QKD) and post-quantum cryptography (PQC). Novel to our system is the dynamic obfuscation of the QKD-PQC sequence of operations, the number of operations, and parameters related to the operations; coupled to the integration of a GPS-free quantum synchronization protocol within the QKD process. We compare the performance and overhead of our QKD-PQC system relative to a standard QKD system with one-time pad encryption, demonstrating that our design can operate in real time with little additional overhead caused by the new security features. Since our system can offer additional defensive strategies against a wide spectrum of practical attacks that undermine deployed QKD, PQC, and certain combinations of these two primitives, we suggest that our design represents one of the most secure communication systems currently available. Given the dynamic nature of its obfuscation attributes, our new system can also be adapted in the field to defeat yet-to-be-discovered practical attacks.

quant-ph↗

arXiv2025

LANTERN: Scalable Distillation of Large Language Models for Job-Person Fit and Explanation

Large language models (LLMs) have achieved strong performance across a wide range of natural language processing tasks. However, deploying LLMs at scale for domain specific applications, such as job-person fit and explanation in job seeking platforms, introduces distinct challenges. At LinkedIn, the job person fit task requires analyzing a candidate's public profile against job requirements to produce both a fit assessment and a detailed explanation. Directly applying open source or finetuned LLMs to this task often fails to yield high quality, actionable feedback due to the complexity of the domain and the need for structured outputs. Moreover, the large size of these models leads to high inference latency and limits scalability, making them unsuitable for online use. To address these challenges, we introduce LANTERN, a novel LLM knowledge distillation framework tailored specifically for job person fit tasks. LANTERN involves modeling over multiple objectives, an encoder model for classification purpose, and a decoder model for explanation purpose. To better distill the knowledge from a strong black box teacher model to multiple downstream models, LANTERN incorporates multi level knowledge distillation that integrates both data and logit level insights. In addition to introducing the knowledge distillation framework, we share our insights on post training techniques and prompt engineering, both of which are crucial for successfully adapting LLMs to domain specific downstream tasks. Extensive experimental results demonstrate that LANTERN significantly improves task specific metrics for both job person fit and explanation. Online evaluations further confirm its effectiveness, showing measurable gains in job seeker engagement, including a 0.24\% increase in apply rate and a 0.28\% increase in qualified applications.

cs.CL↗

arXiv2025

MAUSAM: An Observations-focused assessment of Global AI Weather Prediction Models During the South Asian Monsoon

Accurate weather forecasts are critical for societal planning and disaster preparedness. Yet these forecasts remain challenging to produce and evaluate, especially in regions with sparse observational coverage. Current evaluation of artificial intelligence (AI) weather prediction relies primarily on reanalyses, which can obscure important deficiencies. Here we present MAUSAM (Measuring AI Uncertainty during South Asian Monsoon), an evaluation of seven leading AI-based forecasting systems - FourCastNet, FourCastNet-SFNO, Pangu-Weather, GraphCast, Aurora, AIFS, and GenCast - during the South Asian Monsoon, using ground-based weather stations, rain gauge networks, and geostationary satellite imagery. The AI models demonstrate impressive forecast skill during monsoon across a broad range of variables, ranging from large-scale surface temperature and winds to precipitation, cloud cover, and subseasonal to seasonal eddy statistics, highlighting the strength of data-driven weather prediction. However, the models still exhibit systematic errors at finer scales like the underprediction of extreme precipitation, divergent cyclone tracks, and the mesoscale kinetic energy spectra, highlighting avenues for future improvement. A comparison against observations reveals forecast errors up to 15-45% larger than those relative to reanalysis and traditional forecasts, indicating that reanalysis-centric benchmarks can overstate forecast skill. Of the models assessed, AIFS achieves the most consistent representation of atmospheric variables, with GraphCast and GenCast also showing strong skill. The analysis presents a framework for evaluating AI weather models on regional prediction and highlights both the promise and current limitations of AI weather prediction in data-sparse regions, underscoring the importance of observational evaluation for future operational adoption.

physics.ao-ph↗

arXiv2025

Finetuning AI Foundation Models to Develop Subgrid-Scale Parameterizations: A Case Study on Atmospheric Gravity Waves

Global climate models parameterize a range of atmospheric-oceanic processes like gravity waves, clouds, moist convection, and turbulence that cannot be sufficiently resolved. These subgrid-scale closures for unresolved processes are a leading source of model uncertainty. Here, we present a new approach to developing machine learning parameterizations of small-scale climate processes by fine-tuning a pre-trained AI foundation model (FM). FMs are largely unexplored in climate research. A pre-trained encoder-decoder from a 2.3 billion parameter FM (NASA and IBM Research's Prithvi WxC) -- which contains a latent probabilistic representation of atmospheric evolution -- is fine-tuned (or reused) to create a deep learning parameterization for atmospheric gravity waves (GWs). The parameterization captures GW effects for a coarse-resolution climate model by learning the fluxes from an atmospheric reanalysis with 10 times finer resolution. A comparison of monthly averages and instantaneous evolution with a machine learning model baseline (an Attention U-Net) reveals superior predictive performance of the FM parameterization throughout the atmosphere, even in regions excluded from pre-training. This performance boost is quantified using the Hellinger distance, which is 0.11 for the baseline and 0.06 for the fine-tuned model. Our findings emphasize the versatility and reusability of FMs, which could be used to accomplish a range of atmosphere- and climate-related applications, leading the way for the creation of observations-driven and physically accurate parameterizations for more earth-system processes.

physics.ao-ph↗

arXiv2025

360Brew: A Decoder-only Foundation Model for Personalized Ranking and Recommendation

Ranking and recommendation systems are the foundation for numerous online experiences, ranging from search results to personalized content delivery. These systems have evolved into complex, multilayered architectures that leverage vast datasets and often incorporate thousands of predictive models. The maintenance and enhancement of these models is a labor intensive process that requires extensive feature engineering. This approach not only exacerbates technical debt but also hampers innovation in extending these systems to emerging problem domains. In this report, we present our research to address these challenges by utilizing a large foundation model with a textual interface for ranking and recommendation tasks. We illustrate several key advantages of our approach: (1) a single model can manage multiple predictive tasks involved in ranking and recommendation, (2) decoder models with textual interface due to their comprehension of reasoning capabilities, can generalize to new recommendation surfaces and out-of-domain problems, and (3) by employing natural language interfaces for task definitions and verbalizing member behaviors and their social connections, we eliminate the need for feature engineering and the maintenance of complex directed acyclic graphs of model dependencies. We introduce our research pre-production model, 360Brew V1.0, a 150B parameter, decoder-only model that has been trained and fine-tuned on LinkedIn's data and tasks. This model is capable of solving over 30 predictive tasks across various segments of the LinkedIn platform, achieving performance levels comparable to or exceeding those of current production systems based on offline metrics, without task-specific fine-tuning. Notably, each of these tasks is conventionally addressed by dedicated models that have been developed and maintained over multiple years by teams of a similar or larger size than our own.

cs.IR↗

arXiv2025

How and Where to Translate? The Impact of Translation Strategies in Cross-lingual LLM Prompting

Despite advances in the multilingual capabilities of Large Language Models (LLMs), their performance varies substantially across different languages and tasks. In multilingual retrieval-augmented generation (RAG)-based systems, knowledge bases (KB) are often shared from high-resource languages (such as English) to low-resource ones, resulting in retrieved information from the KB being in a different language than the rest of the context. In such scenarios, two common practices are pre-translation to create a mono-lingual prompt and cross-lingual prompting for direct inference. However, the impact of these choices remains unclear. In this paper, we systematically evaluate the impact of different prompt translation strategies for classification tasks with RAG-enhanced LLMs in multilingual systems. Experimental results show that an optimized prompting strategy can significantly improve knowledge sharing across languages, therefore improve the performance on the downstream classification task. The findings advocate for a broader utilization of multilingual resource sharing and cross-lingual prompt optimization for non-English languages, especially the low-resource ones.

cs.CL↗

arXiv2025

Neutrino oscillation measurements with KamLAND and JUNO in the presence of scalar NSI

Determination of neutrino mass ordering and precision measurement of neutrino oscillation parameters are the foremost goals of the JUNO experiment. Here, we explore the effects of scalar non-standard interactions (sNSI) on the electron anti-neutrino survival probability measured by JUNO. sNSI appear as corrections to the neutrino mass term in the Hamiltonian. We have considered the simplest scenario where there is only one NSI ($η_{ee}$) present in the theory. Our results show that sNSI can have a significant effect on neutrino oscillation probabilities at the medium- and long-baseline reactor experiments. We fit KamLAND data assuming non-zero sNSI in theory and find that {\it estimates of $Δm^2_{21}$ and $θ_{12}$ from KamLAND deviate significantly from their standard best-fit values} if one assumes sNSI in the theory. $η_{ee} \in [-1.0, 1.0]$ is allowed by KamLAND. JUNO cannot constrain sNSI but it can robustly measure $Δm^2_{21}$ and $θ_{12}$ even when they differ widely from their current best-fit values. {\it Our work highlights the necessity of global analysis of constraints on sNSI and standard two-flavour oscillation parameters before arduous three-flavour questions such as neutrino mass ordering or CP violation in their presence are attempted.

hep-ph↗

arXiv2025

Multilingual Information Retrieval with a Monolingual Knowledge Base

Multilingual information retrieval has emerged as powerful tools for expanding knowledge sharing across languages. On the other hand, resources on high quality knowledge base are often scarce and in limited languages, therefore an effective embedding model to transform sentences from different languages into a feature vector space same as the knowledge base language becomes the key ingredient for cross language knowledge sharing, especially to transfer knowledge available in high-resource languages to low-resource ones. In this paper we propose a novel strategy to fine-tune multilingual embedding models with weighted sampling for contrastive learning, enabling multilingual information retrieval with a monolingual knowledge base. We demonstrate that the weighted sampling strategy produces performance gains compared to standard ones by up to 31.03\% in MRR and up to 33.98\% in Recall@3. Additionally, our proposed methodology is language agnostic and applicable for both multilingual and code switching use cases.

cs.CL↗

arXiv2025

Searching for dark matter annihilating into light long-lived mediators from stars inside dwarf spheroidal galaxies

Several astrophysical and cosmological observations suggest the existence of dark matter (DM) through its gravitational effects, yet its nature remains elusive. Despite the lack of DM signals from direct detection experiments, efforts continue to focus on the indirect detection of DM from DM-rich astrophysical objects. Dwarf spheroidal galaxies (dSphs) are among the most promising targets for such searches. In this work, we aim to investigate the expected DM capture rate from the stellar component of ten nearby DM-rich dSphs, assuming that the accumulated DM eventually annihilates into light, long-lived mediators (LLLMs) which decay into gamma rays outside the dSphs. We analyze nearly 16 years of {\it Fermi}-LAT data to search for DM annihilation through LLLMs, and, from the observed stacked flux upper limits, set limits on the DM-nucleon scattering cross section for the case of a generic DM scenario. Additionally, we incorporate the Sommerfeld enhancement (SE) effect into the DM annihilation process assuming scalar DM model, and obtain bounds on the DM-nucleon scattering cross section of $\sim~10^{-36} {\rm cm}^2$ for DM masses around 100 GeV. This allows us to explore an alternative avenue for exploring DM phenomena from dSphs and compare our results with the bounds reported by direct DM detection experiments and other celestial bodies.

astro-ph.HE↗

arXiv2025

AlphaPO: Reward Shape Matters for LLM Alignment

Reinforcement Learning with Human Feedback (RLHF) and its variants have made huge strides toward the effective alignment of large language models (LLMs) to follow instructions and reflect human values. More recently, Direct Alignment Algorithms (DAAs) have emerged in which the reward modeling stage of RLHF is skipped by characterizing the reward directly as a function of the policy being learned. Some popular examples of DAAs include Direct Preference Optimization (DPO) and Simple Preference Optimization (SimPO). These methods often suffer from likelihood displacement, a phenomenon by which the probabilities of preferred responses are often reduced undesirably. In this paper, we argue that, for DAAs the reward (function) shape matters. We introduce \textbf{AlphaPO}, a new DAA method that leverages an $α$-parameter to help change the shape of the reward function beyond the standard log reward. AlphaPO helps maintain fine-grained control over likelihood displacement and over-optimization. Compared to SimPO, one of the best performing DAAs, AlphaPO leads to about 7\% to 10\% relative improvement in alignment performance for the instruct versions of Mistral-7B and Llama3-8B while achieving 15\% to 50\% relative improvement over DPO on the same models. The analysis and results presented highlight the importance of the reward shape and how one can systematically change it to affect training dynamics, as well as improve alignment performance.

cs.CL↗

arXiv2025

On the Robustness of Reward Models for Language Model Alignment

The Bradley-Terry (BT) model is widely practiced in reward modeling for reinforcement learning with human feedback (RLHF). Despite its effectiveness, reward models (RMs) trained with BT model loss are prone to over-optimization, losing generalizability to unseen input distributions. In this paper, we study the cause of over-optimization in RM training and its downstream effects on the RLHF procedure, accentuating the importance of distributional robustness of RMs in unseen data. First, we show that the excessive dispersion of hidden state norms is the main source of over-optimization. Then, we propose batch-wise sum-to-zero regularization (BSR) to enforce zero-centered reward sum per batch, constraining the rewards with extreme magnitudes. We assess the impact of BSR in improving robustness in RMs through four scenarios of over-optimization, where BSR consistently manifests better robustness. Subsequently, we compare the plain BT model and BSR on RLHF training and empirically show that robust RMs better align the policy to the gold preference model. Finally, we apply BSR to high-quality data and models, which surpasses state-of-the-art RMs in the 8B scale by adding more than 5% in complex preference prediction tasks. By conducting RLOO training with 8B RMs, AlpacaEval 2.0 reduces generation length by 40% while adding a 7% increase in win rate, further highlighting that robustness in RMs induces robustness in RLHF training. We release the code, data, and models: https://github.com/LinkedIn-XFACT/RM-Robustness.

cs.CL↗

arXiv2025

Implementation of classical client universal blind quantum computation with 8-state RSP in current architecture

The future of quantum computing architecture is most likely the one in which a large number of clients are either fully classical or have a very limited quantum capability while a very small number of servers having the capability to perform quantum computations and most quantum computational tasks are delegated to these quantum servers. In this architecture, it becomes very crucial that a classical/semi-classical client is able to keep the delegated data/ computation secure against eavesdroppers as well as the server itself, known as the blindness feature. In 2009, A. Broadbent et. al proposed a universal blind quantum computation (UBQC) protocol based on measurement-based quantum computation (MBQC) that enables a semi-classical client to delegate universal quantum computation to a quantum server, interactively and fetch the results while the computation itself remains blind to the server. In this work, we propose an implementation (with examples) of UBQC in the current quantum computing architecture, a fully classical client, a quantum server (IBM Quantum) and the computation does not proceed interactively (projective measurement basis is not decided by previous measurement outcome). We combined UBQC with the 8-state remote state preparation (RSP) protocol, to blindly prepare the initial cluster state, which is an initial resource state in UBQC protocol, to allow a completely classical client to perform delegated blind quantum computation. Such an implementation has already been shown to be secure in a game-based security setting, which is the weakest security model.

cs.CR↗

arXiv2025

Exploring the gluon Sivers function in photon + jet production

We study transverse single spin asymmetries (TSSAs) in back-to-back photon plus jet production in the scattering of unpolarized beams of protons off a transversely polarized proton target as probes of Gluon Sivers Function (GSF). We provide estimates within the region where the imbalance $\vec{q}_{\perp} \equiv \vec{p}_{γ\perp} + \vec{p}_{J\perp}$ between the transverse momenta of the photon and the jet is much smaller than the average of the transverse momenta $\vec{p}_{\perp} \equiv \lvert \vec{p}_{γ\perp} - \vec{p}_{J\perp} \rvert / 2$ i.e. the back-to-back region. The presence of these two scales in the process makes it particularly suitable for applying transverse momentum dependent (TMD) factorization, in contrast to single-scale processes that rely on the generalized parton model approach. We present our preliminary results of the transverse single spin asymmetry for this process at $\sqrt{s} $ = 200 GeV.

hep-ph↗

arXiv2025

Combined Quantum and Post-Quantum Security for Earth-Satellite Channels

Experimental deployment of quantum communication over Earth-satellite channels opens the way to a secure global quantum Internet. In this work, we present results from a real-time prototype quantum key distribution (QKD) system, which entails the development of optical systems including the encoding of entangled photon pairs, the development of transmitters for quantum signaling through an emulated Earth-satellite channel, and the development of quantum-decoding receivers. A unique aspect of our system is the integration of QKD with existing cryptographic methods to ensure quantum-resistant security, even at low-key rates. In addition, we report the use of specially designed error-reconciliation codes that optimize the security versus key-rate trade-off. Our work demonstrates, for the first time, a deployment of the BBM92 protocol that offers both post-quantum security via the advanced encryption standard (AES) and quantum security via an entanglement-based QKD protocol. If either the AES or the QKD is compromised through some adversary attack, our system still delivers state-of-the-art communications secure against future quantum computers.

quant-ph↗

arXiv2024

Efficient user history modeling with amortized inference for deep learning recommendation models

We study user history modeling via Transformer encoders in deep learning recommendation models (DLRM). Such architectures can significantly improve recommendation quality, but usually incur high latency cost necessitating infrastructure upgrades or very small Transformer models. An important part of user history modeling is early fusion of the candidate item and various methods have been studied. We revisit early fusion and compare concatenation of the candidate to each history item against appending it to the end of the list as a separate item. Using the latter method, allows us to reformulate the recently proposed amortized history inference algorithm M-FALCON \cite{zhai2024actions} for the case of DLRM models. We show via experimental results that appending with cross-attention performs on par with concatenation and that amortization significantly reduces inference costs. We conclude with results from deploying this model on the LinkedIn Feed and Ads surfaces, where amortization reduces latency by 30\% compared to non-amortized inference.

cs.LG↗

arXiv2024

WxC-Bench: A Novel Dataset for Weather and Climate Downstream Tasks

High-quality machine learning (ML)-ready datasets play a foundational role in developing new artificial intelligence (AI) models or fine-tuning existing models for scientific applications such as weather and climate analysis. Unfortunately, despite the growing development of new deep learning models for weather and climate, there is a scarcity of curated, pre-processed machine learning (ML)-ready datasets. Curating such high-quality datasets for developing new models is challenging particularly because the modality of the input data varies significantly for different downstream tasks addressing different atmospheric scales (spatial and temporal). Here we introduce WxC-Bench (Weather and Climate Bench), a multi-modal dataset designed to support the development of generalizable AI models for downstream use-cases in weather and climate research. WxC-Bench is designed as a dataset of datasets for developing ML-models for a complex weather and climate system, addressing selected downstream tasks as machine learning phenomenon. WxC-Bench encompasses several atmospheric processes from meso-$β$ (20 - 200 km) scale to synoptic scales (2500 km), such as aviation turbulence, hurricane intensity and track monitoring, weather analog search, gravity wave parameterization, and natural language report generation. We provide a comprehensive description of the dataset and also present a technical validation for baseline analysis. The dataset and code to prepare the ML-ready data have been made publicly available on Hugging Face -- https://huggingface.co/datasets/nasa-impact/WxC-Bench

cs.LG↗

arXiv2024

Machine Learning Global Simulation of Nonlocal Gravity Wave Propagation

Global climate models typically operate at a grid resolution of hundreds of kilometers and fail to resolve atmospheric mesoscale processes, e.g., clouds, precipitation, and gravity waves (GWs). Model representation of these processes and their sources is essential to the global circulation and planetary energy budget, but subgrid scale contributions from these processes are often only approximately represented in models using parameterizations. These parameterizations are subject to approximations and idealizations, which limit their capability and accuracy. The most drastic of these approximations is the "single-column approximation" which completely neglects the horizontal evolution of these processes, resulting in key biases in current climate models. With a focus on atmospheric GWs, we present the first-ever global simulation of atmospheric GW fluxes using machine learning (ML) models trained on the WINDSET dataset to emulate global GW emulation in the atmosphere, as an alternative to traditional single-column parameterizations. Using an Attention U-Net-based architecture trained on globally resolved GW momentum fluxes, we illustrate the importance and effectiveness of global nonlocality, when simulating GWs using data-driven schemes.

physics.ao-ph↗

arXiv2024

DARD: A Multi-Agent Approach for Task-Oriented Dialog Systems

Task-oriented dialogue systems are essential for applications ranging from customer service to personal assistants and are widely used across various industries. However, developing effective multi-domain systems remains a significant challenge due to the complexity of handling diverse user intents, entity types, and domain-specific knowledge across several domains. In this work, we propose DARD (Domain Assigned Response Delegation), a multi-agent conversational system capable of successfully handling multi-domain dialogs. DARD leverages domain-specific agents, orchestrated by a central dialog manager agent. Our extensive experiments compare and utilize various agent modeling approaches, combining the strengths of smaller fine-tuned models (Flan-T5-large & Mistral-7B) with their larger counterparts, Large Language Models (LLMs) (Claude Sonnet 3.0). We provide insights into the strengths and limitations of each approach, highlighting the benefits of our multi-agent framework in terms of flexibility and composability. We evaluate DARD using the well-established MultiWOZ benchmark, achieving state-of-the-art performance by improving the dialogue inform rate by 6.6% and the success rate by 4.1% over the best-performing existing approaches. Additionally, we discuss various annotator discrepancies and issues within the MultiWOZ dataset and its evaluation system.

cs.CL↗