15064 research outputs found
Sort by
The Marks That I Bear: A Theological Anthropology of Queer and Trans BIPOC Tattooing Spaces
No full textNo abstrac
Follow the Money--From Roots to Reality: Encountering the Theology and Ethics of Wealth
No full textNo abstrac
Observation of Fermi Acceleration with Cold Atoms
Get PDFCosmic rays are deemed to be generated by a process known as “Fermi acceleration” in which charged particles scatter against magnetic fluctuations in astrophysical plasmas. The process itself is, however, universal, has both classical and quantum formulations, and is at the basis of dynamical systems with interesting mathematical properties, such as the celebrated Fermi-Ulam model. Despite its effectiveness in accelerating particles, Fermi acceleration has so far eluded unambiguous verifications in laboratory settings. Here, we realize a fully controllable Fermi accelerator by colliding ultracold atoms against engineered movable potential barriers. We demonstrate that our Fermi accelerator, which is only in size, can produce ultracold atomic jets with velocities above 0.5 m/s. Adding dissipation, we also experimentally test Bell’s general argument for the ensuing energy spectra, which is at the basis of any model of cosmic ray acceleration. On the one hand, our Letter effectively opens the window to the use of cold atoms to study phenomena relevant for high energy astrophysics. On the other, the performance of our Fermi accelerator is competitive with those of best-in-class accelerating methods used in quantum technology and quantum colliders, but with substantially simpler implementation no fundamental physics limit
The Sound of Bhojpuri: Screen Media, Language and Politics in Vernacular North India
No full textSince the early 2000s, the rise of Bhojpuri screen media – films, songs, music videos, and more recently, digital media artefacts like short fiction videos – has imparted wide recognizability to the Bhojpuri language in India. Such is the appeal of these Bhojpuri media artefacts that they circulate widely beyond the Bhojpuri-speaking region of India as a larger, vernacular, model of media production, contrasted with those circulated by global media industry centered around the Hindi language, aka Bollywood. Bhojpuri as language, however, continues to seek official recognition as a language from the Indian state, which considers it to be a dialect of the North Indian hegemonic standard, Hindi. Bhojpuri language activists, intellectuals, and media workers are thus situated across a divide where Bhojpuri media’s popularity does not translate into wider, official recognition for the very language that these media objects nominally represent. On the contrary, the subordinate status of the Bhojpuri language as a dialect of Hindi is deepened by the perceived inferiority of Bhojpuri media, considered to be lowbrow, derivative forms by the more established language media industries of India. Bhojpuri thus presents an atypical situation where language and screen media are joined at the hip, where the vernacular ceases to be a mere speech form and language is more than a nominal classificatory category for screen media. Taking Bhojpuri as an atypical, but increasingly routine, paradigm of forging vernacular, ethnolinguistic identity, I ask what makes Bhojpuri a site of investment by screen media creators and language activists alike, and how. The dissertation argues that Bhojpuri screen media and language both distinguish themselves through semiotic processes that enact specifically Bhojpuri perspectives on the world. I unpack three processes at work here: chronotopy, voicing/imaging, and enregisterment. Chronotopy, a Bakhtinian concept denoting the mediation of experience through the ways in which representations link time, space, and personhood, conjures up the figures and scenes of social life that scaffold Bhojpuri screen media and literary writing. Imparting a perspective to these scenes by voicing and imaging them in a way that can be registered and recognized as Bhojpuri by those who receive it, Bhojpuri is made into a recognizable paradigm of doing things in the world in contrast to others. Despite incorporating features and qualities that are not local to the Bhojpuri-speaking region, and are often transnational, Bhojpuri is made into and registered as a local, vernacular category distinct from others that one can belong to. This semiotic enregisterment of Bhojpuri – its establishment as a contrasted and distinct category of persons, speech, images, media, community – as a vernacular way of acting in the world has significance for the social and political life of postcolonial India where the vernacular is increasingly made a force of decolonization. The dissertation is based on ten months of continuous ethnographic fieldwork, supplemented by over six months of phased intermittent fieldwork, in three Indian cities – Delhi, Mumbai, and Varanasi, all major centers of Bhojpuri cultural production. Drawing on this ethnographic research featuring interviews with Bhojpuri media creators, participant observation at institutional and informal sites of media production and language activism, group discussions with Bhojpuri writers and intellectuals, this thesis argues that sounding Bhojpuri is not merely a matter of producing speech in a language – rather, it amounts to the making of situated perspectives for social action. By analyzing Bhojpuri ethnolinguistic identity as a phenomenon whose sites are spread across media and language use, this dissertation contributes to the growing literature on ‘cultural’ media in media studies and to the semiotic analysis of media in anthropology. The dissertation makes the claim that cultural media, such as the practices named Bhojpuri, reveal how the black box of ‘culture’ is made an everyday object of concern
The Masses, Compositions, and Orbits of Planets Around Nearby M Dwarfs
Get PDFWhile thousands of exoplanets have been discovered to date, there are still many open questions with regards to their formation, evolution, and occurrence. These questions are especially difficult to answer for the smallest (and most difficult to observe) planets. As the typical planet signal is inversely proportional to the size of its host star, M dwarfs provide us the best opportunity to study small, Earth-like planets. The radii, masses, and stellar properties of these planetary systems will be necessary in order to understand their compositions, where they formed, their migration history, and even the state of their atmospheres. While photometric surveys such as TESS and Kepler can give us insight into the orbital periods and radii of these planets, follow-up of these systems is necessary in order to measure their masses and orbits. In this dissertation, I highlight several different projects that directly contribute to our understanding of M dwarf planet occurrence, formation, and evolution using the MAROON-X instrument. Firstly, I discuss several different projects in which I used MAROON-X to measure the orbital obliquities of two M dwarf systems (TRAPPIST-1 and LP 267-75) using the Rossiter-McLaughlin effect. Overall, I found that these M dwarfs were aligned with the orbits of their planets, which may imply that small, fully-convective M dwarfs are highly effective at aligning their planets. Later, I describe several projects that utilized MAROON-X’s stability and red wavelength coverage to measure the masses of M dwarf planets. These projects include HUMDRUM (Hunting for M Dwarf Rocky Planets Using MAROON-X), a volume-limited survey I led that measured the masses of nearby M dwarf rocky planets that had transits identified via TESS. Overall, I found that rocky M dwarf planets tend to have Earthlike or slightly sub-Earth densities. However, many of these planets are unlikely to have thick atmospheres, making them difficult to follow-up with atmospheric reconnaissance studies with instruments like JWST
On the Symbiosis of Generative Modeling and Representation Learning
Get PDFGenerative modeling and representation learning are core pillars of modern machine learning and computer vision. In recent years, the field has progressed from analyzing existing visual data to building generative models that can synthesize realistic and diverse visual content. These models offer not only powerful tools for content creation but also a unique perspective on visual understanding—by learning to reconstruct visual structures, they reveal how patterns can be captured, organized, and computationally represented. This thesis investigates the bidirectional relationship between generative modeling and representation learning through two complementary perspectives. The first part of this thesis focuses on enhancing the representation learning capabilities of generative models. We begin by identifying a key limitation in standard architectural designs, specifically, how residual connections in generative models tend to favor high-rank features, which biases learning toward low-level textures rather than semantically meaningful abstractions. To address this, we introduce a decayed residual connection that penalizes the contribution of skip connections, effectively encouraging the model to learn compact, low-rank representations. This design significantly improves both representation quality and generative performance in masked autoencoders and diffusion-based models. However, although diffusion models inherently learn useful representations, obtaining a compact and coherent low-dimensional embedding remains difficult due to the distributed nature of the representation across multiple noise levels and layers. Inspired by classical spectral methods, we propose an efficient distributed spectral clustering algorithm that aggregates features from various stages of the model to form a compact, semantically rich embedding. We further extend our analysis to the generative adversarial network (GAN) framework. Observing that GAN discriminators often learn meaningful features, we introduce a novel representation-aware learning objective along with a capacity-preserving regularization technique. This approach enhances the quality of features learned by the discriminator, yielding improvements that make them useful for downstream semantic tasks. The second part of this thesis examines how learned representations can be used to enhance the quality of generation. We develop a hierarchical generative model that operates in a cascade of semantic spaces, ranging from global structure to fine-grained details, extracted from a pretrained visual encoder. A set of diffusion models is trained to sequentially reconstruct these semantic features using denoising objectives. We demonstrate that a semantic-aware latent representation, such as a 256-dimensional vector from a CLIP encoder, achieves a significantly higher compression ratio than traditional VAE latents, preserving almost all visual information in a 256×256 image. This architecture not only improves sample quality but also accelerates training and outperforms larger models that use more data. Finally, we explore how physics-informed representations can further enhance generation capabilities. By incorporating an autoencoder with a latent bottleneck designed to reflect physical properties—specifically, intrinsic reflectance and lighting—we enable the model to disentangle and manipulate scene properties. This allows for unsupervised generation of albedo maps and realistic image relighting
An Analysis of the Principle of Just Savings and Economic (De)Growth: A Sustainable Median Between Accumulation Stage and Steady State Stage in Relation to Climate Change
Get PDFIn this thesis I argue that climate change challenges John Rawls’s justice as fairness framework by creating a tension between economic growth in the accumulation stage and the Principle of Just Savings, which mandates resource preservation for future generations. Fossil fuel-driven growth exacerbates environmental degradation, perpetuating an unsustainable accumulation stage that undermines intergenerational justice. I propose a hybrid model that integrates the Degrowth Model’s focus on reduced consumption with the Green Transition Model’s emphasis on renewable energy to reconcile domestic and intergenerational justice. By addressing critiques from Hyunseop Kim, Susan Moller Okin, and Martha Nussbaum, the hybrid model extends Rawls’s theory to non-ideal, global contexts, offering a sustainable framework for climate justice that balances economic development with ecological and ethical imperatives in international relations
The Restorative Power of Naps: Sleep-Dependent Consolidation of Generalized Perceptual Learning
Get PDFMemory consolidation is a critical process by which labile learning is stabilized for long-term retention. While a growing body of research supports the role of sleep in consolidating memory, much of this work has focused on rote learning and overnight sleep. Less is known about whether generalized learning — where learners extract abstract patterns that can transfer across novel contexts — also benefits from shorter sleep opportunities, such as naps. Moreover, the mechanisms by which sleep supports such generalization remain unclear. In a series of behavioral and polysomnographic experiments, participants were trained to recognize synthetic speech stimuli in which no words were repeated. These tasks required learners to generalize beyond memorized items and instead acquire abstract acoustic-phonetic patterns. Chapter 2 asks whether a 90-minute nap is sufficient to stabilize generalized learning, compared to remaining awake. These findings suggest that a single nap can consolidate generalized learning, mimicking effects of overnight sleep. Chapter 3 replicates these behavioral findings using a revised design and confirms that participants show performance recovery after a nap. Importantly, this chapter introduces EEG measures and shows that the presence of the EEG cap does not disrupt consolidation. Chapter 4 investigates whether sleep history prior to the nap influences the effectiveness of sleep-based consolidation. Results reveal that recent and cumulative sleep restriction are associated with altered sleep architecture during the nap, and that waking out of slow wave sleep is associated with higher recovery post nap. This suggests that an individual’s sleep history can modulate how effectively sleep consolidates learning. Chapter 5 uses polysomnography to examine which features of nap sleep predict individual differences in memory recovery. Contrary to predictions from some consolidation theories, sleep spindle density was not significantly associated with recovery. Instead, both slow wave sleep (SWS) and REM sleep duration were positively associated with improved post-nap performance, as long as both stages were present during the nap. These findings highlight the importance of both sleep stages and suggest that intact sleep cycles may be necessary to support consolidation of generalized perceptual learning. Together, this work challenges the notion that multiple sleep cycles are required for consolidation, extends our understanding of generalized learning, and underscores the role of both sleep history and specific sleep stages — particularly SWS and REM — in supporting memory stabilization