arXiv subjects

Pritam Sen

Publications and source records attributed to Pritam Sen.

10 recordsLinked to original sources

arXiv2026

Bridging the Gap: A Longitudinal Analysis of Extended Identifiers in the Post-Cookie Era

As third-party cookies fade because of browser restrictions, the online advertising ecosystem is turning to extended identifiers (EIDs) as an alternative. EIDs are persistent user identifiers, such as hashed email addresses, that are employed to link users across domains and devices. This paper presents a 41-month longitudinal study examining EID usage in over 145 million HTTP header bidding requests sent to six major supply-side platforms (SSPs) from 616,539 websites. Our findings show that EIDs are widely used and are becoming increasingly prevalent in the digital advertising ecosystem, reaching 83.76% of studied websites by May 2025. Our analysis of the 18 popular EID providers that account for 99.42% of all transmitted EIDs in our dataset raises concerns about the readiness of EIDs as an alternative to third-party cookie tracking. In terms of accuracy, only one identity provider consistently recognizes and identifies that the visitor is a self-identified bot crawler, and many providers regularly transmit multiple EIDs for the same visitor. We also identify privacy concerns with EIDs, as 12 of the providers create persistent EIDs that can identify the same user across visits, websites, devices, and months. Finally, we found that 16 providers transmit EIDs on EU websites without user consent.

cs.NI↗

arXiv2026

Finite Temperature NLO Corrections in Relativistic Scatterings: Implications for Dark Matter Freeze-In

We study the next-to-leading order (NLO) virtual and thermal corrections to relativistic $2 \rightarrow 2$ scattering processes involving scalar particles in the early Universe thermal plasma. Taking the example of freeze-in production of scalar dark matter pairs through these scatterings, we evaluate the impact of the NLO corrections to the annihilation rate and the dark matter yield. We find that including only thermal mass corrections to a leading order interaction rate can overestimate the reduction in these rates, and the full NLO corrections can modify the DM abundance predictions by $\mathcal{O}(30\%)$. It is also observed that while the virtual NLO effects are larger, the finite temperature NLO corrections to the matrix elements in the relativistic regime can modify the DM abundance by $\mathcal{O}(10\%)$, in comparison to the virtual NLO corrections.

hep-ph↗

arXiv2026

AdFL: In-Browser Federated Learning for Online Advertisement

Since most countries are coming up with online privacy regulations, such as GDPR in the EU, online publishers need to find a balance between revenue from targeted advertisement and user privacy. One way to be able to still show targeted ads, based on user personal and behavioral information, is to employ Federated Learning (FL), which performs distributed learning across users without sharing user raw data with other stakeholders in the publishing ecosystem. This paper presents AdFL, an FL framework that works in the browsers to learn user ad preferences. These preferences are aggregated in a global FL model, which is then used in the browsers to show more relevant ads to users. AdFL can work with any model that uses features available in the browser such as ad viewability, ad click-through, user dwell time on pages, and page content. The AdFL server runs at the publisher and coordinates the learning process for the users who browse pages on the publisher's website. The AdFL prototype does not require the client to install any software, as it is built utilizing standard APIs available on most modern browsers. We built a proof-of-concept model for ad viewability prediction that runs on top of AdFL. We tested AdFL and the model with two non-overlapping datasets from a website with 40K visitors per day. The experiments demonstrate AdFL's feasibility to capture the training information in the browser in a few milliseconds, show that the ad viewability prediction achieves up to 92.59% AUC, and indicate that utilizing differential privacy (DP) to safeguard local model parameters yields adequate performance, with only modest declines in comparison to the non-DP variant.

cs.CR↗

arXiv2026

Inconsistencies in Classification of Online News Articles: A Call for Common Standards in Brand Safety Services

This study examines inconsistencies in the brand safety classifications of online news articles by analyzing ratings from three leading brand safety providers, DoubleVerify, Integral Ad Science, and Oracle. We focus on news content because of its central role in public discourse and the significant financial consequences of unsafe classifications in a sector that is already underserved by digital ad spending. By collecting data from 4,352 news articles on 51 domains, our analysis shows that brand safety services often produce conflicting classifications, with significant discrepancies between providers. These inconsistencies can have harmful consequences for both advertisers and publishers, leading to misplaced advertising spending and revenue losses. This research provides critical insights into the shortcomings of the current brand safety landscape. We argue for a standardized and transparent brand safety system to mitigate the harmful effects of the current system on the digital advertising ecosystem.

cs.CY↗

arXiv2025

An effective field theory for thermal QCD with 2+1 flavours

We write a long-distance effective field theory (EFT) for QCD at finite temperature just below the crossover temperature $T_c$. The low energy constants (LECs) of this EFT are obtained from lattice measurements of the screening mass of pions at two temperatures for $N_f=2+1$ using lattice results obtained at physical values of pion and Kaon masses, and $N_f=2$ where the lattice simulations were performed with a heavier pion mass. The EFT gives good predictions for other static pion properties for $N_f=2$, where lattice results are available. We show the corresponding predictions for $N_f=2+1$, where they are not yet measured. We demonstrate that EFT gives excellent predictions for the phase diagram in $N_f=2+1$. The predictions for the pressure are investigated, and predictions are also given for a Wick-rotated real-time quantity called the kinetic mass.

hep-lat↗

arXiv2025

CryptGNN: Enabling Secure Inference for Graph Neural Networks

We present CryptGNN, a secure and effective inference solution for third-party graph neural network (GNN) models in the cloud, which are accessed by clients as ML as a service (MLaaS). The main novelty of CryptGNN is its secure message passing and feature transformation layers using distributed secure multi-party computation (SMPC) techniques. CryptGNN protects the client's input data and graph structure from the cloud provider and the third-party model owner, and it protects the model parameters from the cloud provider and the clients. CryptGNN works with any number of SMPC parties, does not require a trusted server, and is provably secure even if P-1 out of P parties in the cloud collude. Theoretical analysis and empirical experiments demonstrate the security and efficiency of CryptGNN.

cs.CR↗

arXiv2025

Thermal corrections to dark matter annihilation with real photon emission/absorption

The dark matter relic density is being increasingly precisely measured. This relic density is theoretically determined by a Boltzmann equation which computes the dark matter distribution according to the (thermally averaged) cross section for annihilation/production of dark matter within a given model. We present here the complete higher order thermal corrections, calculated using real time thermal field theory, to the cross section for the annihilation of dark matter into Standard Model fermions via charged scalars: $\chi \chi \to f \overline{f}$ and $\chi \chi \to f \overline{f} (\gamma)$. The latter process includes real photon emission into, and absorption from, the heat bath at temperature $T$. We use the Grammer and Yennie technique to separate the soft infra-red divergences, which greatly simplifies the calculation. We show explicitly the cancellation of both soft and collinear divergences between real and virtual contributions at next-to-leading order (NLO) in the thermal field theory and remark on the non-trivial nature of the collinear divergences when the thermal contribution from fermions is considered. We present the leading thermal contributions to order ${\cal{O}}(\alpha T^2)$ for the case when the dark matter particle is a Majorana or Dirac type fermion. In the former case, both the leading order (LO) and NLO cross sections are helicity suppressed, while neither is suppressed in the Dirac case. It is interesting that the ratio of NLO to LO cross sections is the same for both Majorana and Dirac type dark matter.

hep-ph↗

arXiv2024

NLO thermal corrections to dark matter annihilation cross sections: a novel approach

The dark matter relic density has been increasingly accurately measured by successive generations of experiments. The Boltzmann equation determines the yields using the dark matter annihilation cross section as one of the inputs; the accurate computation of the latter including thermal contributions thus assumes importance. We report here the next-to-leading order (NLO) thermal corrections to the cross sections for (Majorana) dark matter annihilation to standard model fermions: $\chi \chi \to f \overline{f}$, via charged scalars. We use a novel approach, utilising the technique of Grammer and Yennie, extended to thermal field theories, where the cancellation of soft infra-red divergences occurs naturally. We present the NLO thermal cross sections in full detail for both the relativistic case as well as in the non-relativistic limit. Our independent calculation verifies earlier results where the leading contribution at order ${\cal{O}}(T^2)$ was shown to be proportional to the square of the fermion mass in the non-relativistic limit, just as at leading order. We find that the ${\cal{O}}(T^4)$ contributions have the same dependence on the fermion mass as well.

hep-ph↗

arXiv2019

Infrared finiteness of theories with bino-like dark matter at finite temperature

Models incorporating moderately heavy dark matter (DM) typically need charged (scalar) fields to establish admissible relic densities. Since the DM freezes out at an early epoch, thermal corrections to the cross sections can be important. In a companion paper [arXiv:1812.04247v2] we established that the infrared (IR) divergences accruing from scalar-photon interactions cancel to all orders in perturbation theory. The corresponding infrared finiteness of thermal fermionic QED has already been established. Here, we study the IR behaviour at finite temperatures, of a theory of dark matter interacting with charged scalars and fermions, which potentially contains both both linear and sub-leading logarithmic divergences. We prove that the theory is IR-finite to all orders with the divergences cancelling when both absorption and emission of photons from and into the heat bath are taken into account. While 4-point interaction terms are known to be IR finite, their inclusion leads to a neat exponentiation. The calculation follows closely the technique used for the scalar finite temperature theory.

hep-th↗

arXiv2019

Infrared finiteness of a thermal theory of scalar electrodynamics to all orders

Models explaining dark matter typically include interactions with charged scalar and fermion fields. The Infra-Red (IR) finiteness of thermal field theories of charged fermions (fermionic QED) has been proven to all orders in perturbation theory. Here we reexamine the IR behaviour of charged scalar theories at finite temperature. Using the method of Grammer and Yennie, we identify and factorise the infra-red divergences to all orders in perturbation theory. The inclusion of IR finite pieces arising from the 4-point interaction terms of scalars with photon fields is key to the exponentiation. We use this in a companion paper to prove the IR finiteness of the corresponding thermal theory which is of relevance in dark matter calculations.

hep-ph↗