Search

With an evergrowing number of LLMs reporting superlative performance for English, their ability to perform equitably for different dialects of English ($\textit{i.e.}$, dialect robustness) needs to be ascertained. Specifically, we use English language (US English or Indian English) conversations between humans who play the word-guessing game of 'taboo'. We formulate two evaluative tasks: target word prediction (TWP) ($\textit{i.e.}$, predict the masked target word in a conversation) and target word selection (TWS) ($\textit{i.e.}$, select the most likely masked target word in a conversation, from among a set of candidate words). Extending MD3, an existing dialectic dataset of taboo-playing conversations, we introduce M-MD3, a target-word-masked version of MD3 with the en-US and en-IN subsets. We create two subsets: en-MV (where en-US is transformed to include dialectal information) and en-TR (where dialectal information is removed from en-IN). We evaluate one open-source (Llama3) and two closed-source (GPT-4/3.5) LLMs. LLMs perform significantly better for US English than Indian English for both TWP and TWS tasks, for all settings, exhibiting marginalisation against the Indian dialect of English. While GPT-based models perform the best, the comparatively smaller models work more equitably after fine-tuning. Our error analysis shows that the LLMs can understand the dialect better after fine-tuning using dialectal data. Our evaluation methodology exhibits a novel way to examine attributes of language models using pre-existing dialogue datasets., Comment: 12 pages, 3 figures, 7 tables

The pervasive influence of social biases in language data has sparked the need for benchmark datasets that capture and evaluate these biases in Large Language Models (LLMs). Existing efforts predominantly focus on English language and the Western context, leaving a void for a reliable dataset that encapsulates India's unique socio-cultural nuances. To bridge this gap, we introduce IndiBias, a comprehensive benchmarking dataset designed specifically for evaluating social biases in the Indian context. We filter and translate the existing CrowS-Pairs dataset to create a benchmark dataset suited to the Indian context in Hindi language. Additionally, we leverage LLMs including ChatGPT and InstructGPT to augment our dataset with diverse societal biases and stereotypes prevalent in India. The included bias dimensions encompass gender, religion, caste, age, region, physical appearance, and occupation. We also build a resource to address intersectional biases along three intersectional dimensions. Our dataset contains 800 sentence pairs and 300 tuples for bias measurement across different demographics. The dataset is available in English and Hindi, providing a size comparable to existing benchmark datasets. Furthermore, using IndiBias we compare ten different language models on multiple bias measurement metrics. We observed that the language models exhibit more bias across a majority of the intersectional groups.

Investigating the polarization properties of light in the mid-infrared (mid-IR) spectrum is crucial for molecular sensing, biomedical diagnostics, and IR imaging system technologies. Traditional methods, limited by bulky size and intricate fabrication, utilize large rotating optics for full Stokes polarization detection, impeding miniaturization and accuracy. Van der Waals materials (vdW) based devices can address these challenges due to their lithography-free fabrication, ease of integration with chip-scale platforms and room-temperature operation. This study introduces a chip-integrated polarimeter device leveraging the in-plane biaxial hyperbolic vdW crystal properties for mid-infrared light manipulation. The spatial division measurement scheme incorporates six meticulously designed linear and circular polarization filters, achieving high extinction ratios exceeding 30 dB and transmittance surpassing 50%, with fabrication tolerance of film thickness up to 100 nm. The proposed device represents a significant advancement in polarimetric detection, providing a compact, cost-effective solution and opens new avenues for on-chip mid-IR polarimetric detection in next-generation ultra-compact optical systems.

The heightened sensitivity observed in non-Hermitian systems at exceptional points (EPs) has garnered significant attention. Typical EP sensor implementations rely on precise measurements of spectra and importantly, for real time sensing measurements, the EP condition ceases to hold as the perturbation increases over time, thereby preventing the use of high sensitivity at the EP point. In this work, we present an new approach to EP sensing which goes beyond these two traditional constraints. Firstly, instead of measuring the spectra, our scheme of EP based sensing is based on the observation of decay length of the optical mode in finite size gratings, which is validated via coupled mode theory as well as full wave electrodynamic simulations. Secondly, for larger perturbation strengths, the EP is spectrally shifted instead of being destroyed -- this spectral shift of the EP is calibrated and using this look-up table, we propose continuous real time detection by varying the excitation laser wavelength. As a proof of principle of our technique, we present an application to the sensing of coronavirus particles, which shows unprecedented limit of detection. These findings will contribute to the expanding field of exceptional point based sensing technologies for real time applications beyond spectral measurements.

In-plane Hyperbolic Phonon polaritons (HPhPs) are quasiparticles formed via coupling of photons and optical phonons in in-plane hyperbolic materials and offer unique applications in sensing, thermal emitters and high resolution imaging. However, the large momentum mismatch between photons and these in-plane HPhPs has restricted their technological potential as most experimental demonstration rely on sophisticated and expensive near field detection schemes. In this work, using the example of $\alpha-$MoO$_3$, we demonstrate that by constructing photonic hypercrystals of this material, one can not only excite these in-plane HPhPs in the far field but also tune the far field response via twisting the hypercrystal lattice with respect the lattice of $\alpha-$MoO$_3$. Our findings will pave the way for the development of practical in-plane HPhP devices as well as provide access to new fundamental physics of such materials via conventional and well developed far field measurement techniques.

We highlight the STAR experiment's hard probes results, including jets and heavy flavor production in heavy-ion collisions to study the properties of the quark-gluon plasma. Various jet-substructure observables in proton-proton collisions are presented to explore both perturbative and non-perturbative regimes of quantum chromodynamics. Finally, we discuss the STAR experiment's hard probes physics program for ongoing data taking., Comment: Hard Probes 2023 Conference, Aschaffenburg, Germany, STAR highlights plenary talk on hard probes, 10 pages, 9 figures

Enhancing practical low light raw images is a difficult task due to severe noise and color distortions from short exposure time and limited illumination. Despite the success of existing Convolutional Neural Network (CNN) based methods, their performance is not adaptable to different camera domains. In addition, such methods also require large datasets with short-exposure and corresponding long-exposure ground truth raw images for each camera domain, which is tedious to compile. To address this issue, we present a novel few-shot domain adaptation method to utilize the existing source camera labeled data with few labeled samples from the target camera to improve the target domain's enhancement quality in extreme low-light imaging. Our experiments show that only ten or fewer labeled samples from the target camera domain are sufficient to achieve similar or better enhancement performance than training a model with a large labeled target camera dataset. To support research in this direction, we also present a new low-light raw image dataset captured with a Nikon camera, comprising short-exposure and their corresponding long-exposure ground truth images., Comment: BMVC 2021 Best Student Paper Award (Runner-Up). Project Page: https://val.cds.iisc.ac.in/HDR/BMVC21/index.html

Strong coupling of excitons to optical cavity modes is of immense importance to understanding the fundamental physics of quantum electrodynamics at the nanoscale as well as for practical applications in quantum information technologies. There have been several attempts at achieving strong coupling between excitons in two dimensional semiconductors such as transition metal dichalcogenides (TMDCs) and photonic quasi-bound states in the continuum (BICs). We identify two gaps in the platforms for achieving strong coupling between TMDC excitons and photonic quasi-BICs: firstly, in the studies so far, different cavity architectures have been employed for coupling to different TMDCs. This would mean that typically, the fabrication process flow for the cavities will need to be modified as one moves from one TMDC to the other, which can limit the technological progress in the field. Secondly, there has been no discussion of the impact of fabrication imperfections in the studies on strong coupling of these subsystems so far. In this work, we address these two questions by optimizing a cavity with the same architecture which can couple to the four typical TMDCs (MoS$_2$, WS$_2$, MoSe$_2$, WSe$_2$) and perform a detailed investigation on the fabrication tolerance of the associated photonic quasi-BICs and their impact on strong coupling.

With the rise of online hate speech, automatic detection of Hate Speech, Offensive texts as a natural language processing task is getting popular. However, very little research has been done to detect unintended social bias from these toxic language datasets. This paper introduces a new dataset ToxicBias curated from the existing dataset of Kaggle competition named "Jigsaw Unintended Bias in Toxicity Classification". We aim to detect social biases, their categories, and targeted groups. The dataset contains instances annotated for five different bias categories, viz., gender, race/ethnicity, religion, political, and LGBTQ. We train transformer-based models using our curated datasets and report baseline performance for bias identification, target generation, and bias implications. Model biases and their mitigation are also discussed in detail. Our study motivates a systematic extraction of social bias data from toxic language datasets. All the codes and dataset used for experiments in this work are publicly available

The higher-order cumulants of net-proton number, net-charge, and net-strangeness multiplicity distributions are widely studied to search for the quantum-chromodynamics critical point and extract the chemical freeze-out parameters in heavy-ion collisions. In this context, the event-by-event fluctuations of the net-strangeness multiplicity distributions play important roles in extracting the chemical freeze-out parameter in the strangeness sector. Due to having difficulties in detecting all strange hadrons event by event, the kaon ($K$) and lambda ($\Lambda$) particles serve as a proxy for the strangeness-related observables in heavy-ion collisions. We have studied the net-$K$, net-$\Lambda$, and net-($K$ + $\Lambda$) multiplicity distributions and calculated their different order of cumulants using the ultrarelativistic quantum molecular dynamics model and hadron resonance gas calculation. To adequately account for the net-strangeness cumulants, it has been found that the inclusion of resonance decay contributions in $K$ and $\Lambda$ is necessary., Comment: The title has been modified. Updated following the published version

The discovery and characterization of hot and dense QCD matter, known as Quark Gluon Plasma (QGP), remains the most international collaborative effort and synergy between theorists and experimentalists in modern nuclear physics to date. The experimentalists around the world not only collect an unprecedented amount of data in heavy-ion collisions, at Relativistic Heavy Ion Collider (RHIC), at Brookhaven National Laboratory (BNL) in New York, USA, and the Large Hadron Collider (LHC), at CERN in Geneva, Switzerland but also analyze these data to unravel the mystery of this new phase of matter that filled a few microseconds old universe, just after the Big Bang. In the meantime, advancements in theoretical works and computing capability extend our wisdom about the hot-dense QCD matter and its dynamics through mathematical equations. The exchange of ideas between experimentalists and theoreticians is crucial for the progress of our knowledge. The motivation of this first conference named "HOT QCD Matter 2022" is to bring the community together to have a discourse on this topic. In this article, there are 36 sections discussing various topics in the field of relativistic heavy-ion collisions and related phenomena that cover a snapshot of the current experimental observations and theoretical progress. This article begins with the theoretical overview of relativistic spin-hydrodynamics in the presence of the external magnetic field, followed by the Lattice QCD results on heavy quarks in QGP, and finally, it ends with an overview of experiment results., Comment: Compilation of the contributions (148 pages) as presented in the `Hot QCD Matter 2022 conference', held from May 12 to 14, 2022, jointly organized by IIT Goa & Goa University, Goa, India

Movies reflect society and also hold power to transform opinions. Social biases and stereotypes present in movies can cause extensive damage due to their reach. These biases are not always found to be the need of storyline but can creep in as the author's bias. Movie production houses would prefer to ascertain that the bias present in a script is the story's demand. Today, when deep learning models can give human-level accuracy in multiple tasks, having an AI solution to identify the biases present in the script at the writing stage can help them avoid the inconvenience of stalled release, lawsuits, etc. Since AI solutions are data intensive and there exists no domain specific data to address the problem of biases in scripts, we introduce a new dataset of movie scripts that are annotated for identity bias. The dataset contains dialogue turns annotated for (i) bias labels for seven categories, viz., gender, race/ethnicity, religion, age, occupation, LGBTQ, and other, which contains biases like body shaming, personality bias, etc. (ii) labels for sensitivity, stereotype, sentiment, emotion, emotion intensity, (iii) all labels annotated with context awareness, (iv) target groups and reason for bias labels and (v) expert-driven group-validation process for high quality annotations. We also report various baseline performances for bias identification and category detection on our dataset.

These proceedings discuss recent jet measurements by the STAR experiment at RHIC to study jet substructure in p+p and jet quenching in Au+Au collisions at $\sqrt{s_\mathrm{NN}}$ = 200 GeV. Furthermore, STAR's future plans for precision jet measurements with the upcoming data-taking periods in 2023-2025 are presented., Comment: ICNFP2021, Crete, Greece. 8 figures

Integration of conventional mid to long-wavelength infrared polarizers with chip-scale platforms is restricted by their bulky size and complex fabrication. Van der Waals materials based polarizer can address these challenges due to its non-lithographic fabrication, ease of integration with chip-scale platforms, and room temperature operation. In the present work, mid-IR optical response of the sub-wavelength thin films of $\alpha$-MoO$_3$ is investigated for application towards high temperature mid-IR transmission and reflection type thin film polarizer. To our knowledge, this is the first report of above room temperature mid-IR optical response of $\alpha$-MoO$_3$ to determine the thermal stability of the proposed device. We find that our $\alpha$-MoO$_3$ based polarizer retains high extinction ratio with peak value exceeding 10 dB, up to a temperature of 140$^{\circ}$C. We explain our experimental findings by natural in-plane hyperbolic anisotropy of $\alpha$-MoO$_3$ in the mid-IR, high temperature X-ray diffraction and Raman spectroscopic measurements. This work opens up new avenues for naturally in-plane hyperbolic van der Waals thin-films to realize sub-wavelength IR optical components without lithographic constraints., Comment: Corrected affiliation

We present the semi-inclusive measurement of charged jets recoiling from direct-photon and $\pi^{0}$ triggers in central Au+Au collisions at $\sqrt{s_{\mathrm {NN}}}$ = 200 GeV, using a dataset with integrated luminosity 13 $\mathrm {nb^{-1}}$ recorded by the STAR experiment in 2014. The photon and $\pi^{0}$ triggers are selected within transverse energy ($\rm E_\mathrm{T}^\mathrm{trig}$) between 9 GeV and 20 GeV. Charged jets are reconstructed with the anti-$k_{\mathrm{T}}$ algorithm with resolution parameters R = 0.2 and 0.5. A Mixed-Event technique developed previously by STAR is used to correct the recoil jet yield for uncorrelated background, enabling recoil jet measurements over a broad $\rm p_{T,jet}$ range. We report fully corrected charged-jet yields recoiling from direct-photon and $\pi^{0}$ triggers for the above two jet radii and also discuss the jet R dependence of in-medium parton energy loss at the top RHIC energy., Comment: 5 pages, 4 figures, Hard Probes 2020

In this article, there are 18 sections discussing various current topics in the field of relativistic heavy-ion collisions and related phenomena, which will serve as a snapshot of the current state of the art. Section 1 reviews experimental results of some recent light-flavored particle production data from ALICE collaboration. Other sections are mostly theoretical in nature. Very strong but transient magnetic field created in relativistic heavy-ion collisions could have important observational consequences. This has generated a lot of theoretical activity in the last decade. Sections 2, 7, 9, 10 and 11 deal with the effects of the magnetic field on the properties of the QCD matter. There are several unanswered questions about the QCD phase diagram. Sections 3, 11 and 18 discuss various aspects of the QCD phase diagram and phase transitions. Recent years have witnessed interesting developments in foundational aspects of hydrodynamics and their application to heavy-ion collisions. Sections 12, 15, 16 and 17 of this article probe some aspects of this exciting field. Transport coefficients together with their temperature- and density-dependence, are essential inputs in hydrodynamical calculations. Sections 5, 8 and 14 deal with calculation/estimation of various transport coefficients (shear and bulk viscosity, thermal conductivity, relaxation times, etc.) of quark matter and hadronic matter. Sections 4, 6 and 13 deals with interesting new developments in the field. Section 4 discusses color dipole gluon distribution function at small transverse momentum in the form of a series of Bells polynomials. Section 6 discusses the properties of Higgs boson in the quark gluon plasma using Higgs-quark interaction. Section 13 discusses modification of coalescence model to incorporate viscous corrections and application of this model., Comment: 109 pages, 49 captioned figures, compilation of the contributions as presented in the `Workshop on Dynamics of QCD Matter', 15th to 17th August 2019, NISER Bhubaneswar, India, published version

Recently, in-plane biaxial hyperbolicity has been observed in $\alpha$-MoO${_3}$ --a van der Waal crystal-- in the mid-infrared frequency regime. Here, we present a comprehensive theoretical analysis of thin film $\alpha$-MoO${_3}$ for application to two mid-IR photonic devices -- a polarizer and a waveplate. We show the possibility of a significant reduction in the device footprint while maintaining an enormous extinction ratio from $\alpha$-MoO${_3}$ based polarizers in comparison with that of conventional polarizers. Secondly, we carry out device optimization of $\alpha$-MoO${_3}$ based waveplates with subwavelength thickness. We explain our results using natural in-plane hyperbolicity of $\alpha$-MoO${_3}$ via analytical and full wave simulations. This work will build a foundation for miniaturization of mid-infrared photonic devices by exploiting the optical anisotropy of $\alpha$-MoO${_3}$.

We studied the centrality selection effect on cumulants (up to fourth order) and the cumulants ratios of net-proton multiplicity distributions in Au+Au collisions at $\sqrt{s_{\mathrm{NN}}}$ = 7.7, 19.6 and 200 GeV from UrQMD model. The net-proton cumulants are calculated with collision centralities by using charged particle multiplicity from different pesudorapidity ($\eta$) region. By comparing the results from various collision centralities, we found that the autocorrelation effects are not significant in the results with collision centralities "refmult-3" and "refmult-2", which are using mid-rapidity charged particles but excluding (anti-)protons and analysis region, respectively. Furthermore, due to the contributions of spectator protons, we observed poor centrality resolution when using charged particles at forward $\eta$ region at low energies. This work can serve as a baseline for centrality selection of future fluctuations analysis in relativistic heavy-ion collisions., Comment: 11 pages and 12 figures, Minor update on Fig. 5, 6, 12 and references

We report the first measurement of fully corrected semi-inclusive charged recoil jets for both direct-photon ($\gamma_{dir}$) and $\pi^{0}$ triggers, within two trigger transverse energy ranges between 9 GeV and 15 GeV, in central Au+Au collisions at $\sqrt {s_{\rm NN}}$ = 200 GeV using data with an integrated luminosity of 13 $\rm nb^{-1}$ collected by the STAR experiment. A comparison between $\pi^{0}$-triggered recoil jets in p+p collisions and PYTHIA is discussed. A comparison is also presented between the suppression of $\gamma_{dir}$- and $\pi^{0}$-triggered recoil jets in central Au+Au collisions with respect to their corresponding PYTHIA references. A strong and similar level of suppression is observed in recoil-jet yields as a function of jet transverse momentum for $\gamma_{dir}$+jet and $\pi^{0}$+jet., Comment: Hard Probes 2018, 6 figures

We present two measurements related to jet production in p+p and central Au+Au collisions at $\sqrt{{s}_{NN}}$=200 GeV. Firstly, a study of semi-inclusive charged recoil jets coincident with high-$\rm{p_{T}}$ direct-photon and neutral pions, and secondly, the hadron correlations with respect to reconstructed di-jet triggers. Indication of medium effects is observed by comparing $\gamma$+jet and $\pi^{0}$+jet measurements. The di-jet+hadron study shows signs that the medium-induced modifications of an imbalanced set of di-jets with "hard cores" primarily affect the recoil jet., Comment: 4 pages, 3 Figures, proceedings for the Quark Matter 2017

Hard probes created through large momentum transfers are used to study the properties of QCD matter created in heavy-ion collisions, by comparing the measurements to those in p+p collisions. Jets, and the "quenching" or suppression of jets in the medium created in heavy-ion collisions, are studied through various different observables. We present the most recent measurements from $\sqrt{s_{NN}}$ = 200 GeV Au+Au collisions, with p+p collisions as the reference, by the STAR Collaboration. The observables are semi-inclusive charged jets and di-jet transverse momentum imbalance. Additionally, correlation measurements of direct photon-hadron and neutral pion-hadron are presented and discussed., Comment: 5 pages, 4 figures, ICHEP 2016, Chicago, USA

We report the results of $\gamma_{\rm dir}$ and $\pi^{0}$-hadron azimuthal correlations as a measure of the away-side jet-like correlated yields in central Au+Au and p+p collisions at $\sqrt{s_{_{NN}}}$=200 GeV in the STAR experiment. The charged-hadron per-trigger yields at mid-rapidity with respect to high-$p_{T}$ $\gamma_{\rm dir}$ and $\pi^{0}$ in central Au+Au collisions are compared with p+p collisions. Within uncertainties, the same $z_{T}$ (~$p_{T}^{\mathrm{assoc}}$ / $p_{T}^{\mathrm{trig}}$) dependence of the suppression is observed for $\gamma_{\rm dir}$ and $\pi^{0}$ - triggers. The results are compared with energy-loss model predictions. The $\gamma-\rm jet$ measurements can provide further understanding on the redistribution of in-medium energy loss. Ongoing $\gamma-\rm jet$ studies in the STAR experiment are also discussed., Comment: 4 pages, 5 figures, Hot Quarks 2016 conference proceedings

Susceptibilities of conserved quantities, such as baryon number, strangeness and electric charge are sensitive to the onset of quantum chromodynamics (QCD) phase transition and are expected to provide information on the matter produced in heavy-ion collision experiments. A comprehensive study of the second-order diagonal susceptibilities and cross correlations has been made within a thermal model approach of the hadron resonance gas (HRG) model as well as with a hadronic transport model, UrQMD. We perform a detailed analysis of the effect of detector acceptances and choice of particle species in the experimental measurements of the susceptibilities for heavy-ion collisions corresponding to \sNN = 4 GeV to 200 GeV. The transverse momentum cutoff dependence of suitably normalised susceptibilities are proposed as useful observables to probe the properties of the medium at freezeout., Comment: 16 pages, 11 figures

We report new results of $\gamma_{dir}$+hadron and $\pi^{0}$+hadron azimuthal correlations as a measure of the away-side jet-like correlated yields in central Au+Au and p+p collisions at $\sqrt{s_{NN}}$ = 200 GeV in the STAR experiment from years 2011 and 2009 of data taking, respectively. The charged-hadron per-trigger yields at mid-rapidity $(|\eta| < 1)$ and for transverse momenta $p_{T}^{assoc} > 1.2$ ~GeV/$c$ associated with $\gamma_{dir}$~ and $\pi^{0}$~(for triggers $|\eta| < $0.9, 12 $< p_{T}^{trig} <$ 20 GeV/$c$) in central Au+Au collisions are compared with p+p collisions. The $z_{T}$ ($= \frac{p_{T}^{assoc}}{p_{T}^{trig}}$) dependence, now extending down to $z_{T}$=0.1, of the suppression of the away-side associated yields is presented. We observed that the suppression of away-side associated yields tends to vanish at lower \zT. The dependence of the suppression on both $p_{T}^{assoc}$ and $p_{T}^{trig}$ is also discussed. Finally, these results are compared with various model predictions., Comment: Proceedings of the Quark Matter 2015 conference, 4 pages, 5 figures