This is a experimental project. Feel free to send feedback!

Thesis Tide

Thesis Tide ranks papers based on their relevance to the fields, with the goal of making it easier to find the most relevant papers. It uses AI to analyze the content of papers and rank them!

Constructing Uncertainty Sets for Robust Risk Measures: A Composition of $φ$ -Divergences Approach to Combat Tail Uncertainty

Risk measures, which typically evaluate the impact of extreme losses, are highly sensitive to misspecification in the tails. This paper studies a robust optimization approach to combat tail uncertaint...

Useful Fields:

The article presents a novel robust optimization framework that directly addresses tail uncertainty in risk measures, which is a critical aspect in fields like finance and insurance. The methodology it proposes using $φ$-divergences is not only innovative but also pragmatic, offering a computationally feasible means of addressing these uncertainties. This enhancement in methodology has significant implications for real-world applications where precision in risk assessment is paramount.

Physics-informed reduced order model with conditional neural fields

This study presents the conditional neural fields for reduced-order modeling (CNF-ROM) framework to approximate solutions of parametrized partial differential equations (PDEs). The approach combines a...

Useful Fields:

The article presents an innovative approach that integrates physics-informed learning with neural network frameworks to solve parametrized PDEs. The methodology showcases novel applications of neural ODEs and automatic differentiation, thus representing a significant advance in reduced-order modeling. The rigorous validation against established solutions indicates methodological robustness, enhancing its potential impact in related fields.

LIAR: Leveraging Alignment (Best-of-N) to Jailbreak LLMs in Seconds

Many existing jailbreak techniques rely on solving discrete combinatorial optimization, while more recent approaches involve training LLMs to generate multiple adversarial prompts. However, both appro...

Useful Fields:

The article introduces a novel approach to jailbreaking large language models (LLMs) using alignment strategies, significantly improving efficiency and effectiveness. Its theoretical foundation and experimental validation demonstrate both novelty and methodological rigor. The implications for both AI safety and adversarial strategies make this research timely and impactful, especially given the ongoing discussions surrounding model alignments and safety in AI.

Dimensionality reduction for closed-loop quantum gate calibration

Numerical gate design typically makes use of high-dimensional parameterizations enabling sophisticated, highly expressive control pulses. Developing efficient experimental calibration methods for such...

Useful Fields:

The article introduces a novel method for dimensionality reduction specifically tailored to quantum gate calibration, which addresses a critical challenge in quantum control. The approach's systematic nature and its applicability to both amplitude and coherent error robustness enhance its significance. While the concept of dimensionality reduction is not new, its application within the context of quantum gates is relatively innovative and relevant, making it a potential cornerstone for future methodologies in the field.

Profinite rigidity witnessed by Dehn fillings of cusped hyperbolic 3-manifolds

Any profinite isomorphism between two cusped finite-volume hyperbolic 3-manifolds carries profinite isomorphisms between their Dehn fillings. With this observation, we prove that some cusped finite-vo...

Useful Fields:

The article introduces a novel approach to understanding profinite rigidity in hyperbolic 3-manifolds through Dehn fillings, presenting significant findings relevant to the topology of 3-manifolds. The methodological rigor is notable, particularly in the examination of characterizing slopes, which offers a deeper insight into the structure of hyperbolic knots. This work has the potential to inspire future research in low-dimensional topology and geometric group theory, as it addresses foundational aspects of rigidity in 3-manifolds.

MC3: Memory Contention based Covert Channel Communication on Shared DRAM System-on-Chips

Shared-memory system-on-chips (SM-SoC) are ubiquitously employed by a wide-range of mobile computing platforms, including edge/IoT devices, autonomous systems and smartphones. In SM-SoCs, system-wide ...

Useful Fields:

The article presents a novel method for exploiting vulnerabilities in shared-memory architectures in SoCs, which is particularly relevant given the increasing reliance on such systems in mobile and IoT devices. The methodological rigor is evident in its experimental validation across real-world hardware, which strengthens its implications for both security analysis and system design. This work could inspire further research into both defensive strategies against such attacks and advancements in SoC architectures.

A simple non-parametric reconstruction of parton distributions from limited Fourier information

Some calculations of parton distributions from first principles only give access to a limited range of Fourier modes of the function to reconstruct. We present a physically motivated procedure to regu...

Useful Fields:

The article presents a novel approach to reconstructing parton distributions, addressing a common limitation in current methodologies. The proposed use of a Gaussian process adds a valuable Bayesian perspective to the problem, which is likely to enhance the robustness and reliability of the results, making it a significant contribution to theoretical particle physics. The emphasis on interpretability and efficiency also makes the methodology applicable in various scenarios, potentially influencing future research in the field.

Partial Wavefunction Collapse Under Repeated Weak Measurement of a non-Conserved Observable

Two hallmarks of quantum non-demolition (QND) measurement are the ensemble-level conservation of the expectation value of the measured observable $A$ and the eventual, inevitable collapse of t...

Useful Fields:

The article presents innovative findings on quantum measurements, specifically relating to the complex interplay between non-conserved observables and their measurement dynamics. This exploration of non-QND measurements contributes significantly to our understanding of quantum mechanics, introducing a novel concept of conservation in the context of secondary observables. Its rigorous methodological approach, clear theoretical implications, and practical examples demonstrate strong potential for advancing theoretical and experimental research in quantum mechanics and related areas.

BEExformer: A Fast Inferencing Transformer Architecture via Binarization with Multiple Early Exits

Large Language Models (LLMs) based on transformers achieve cutting-edge results on a variety of applications. However, their enormous size and processing requirements make deployment on devices with c...

Useful Fields:

The proposed BEExformer architecture addresses key challenges in deploying large language models (LLMs) on resource-constrained devices, offering significant improvements in both efficiency and accuracy. Its innovative combination of binarization and early exit mechanisms presents a novel approach that could influence future research in transformer architectures and efficiency optimizations. The rigorous evaluation on the GLUE dataset and the clear performance benefits indicate both methodological robustness and practical applicability.

Rectangular recurrence relations in $\mathfrak{gl}_{n}$ and $\mathfrak{o}_{2n+1}$ invariant integrable models

A new method is introduced to derive general recurrence relations for off-shell Bethe vectors in quantum integrable models with either type $\mathfrak{gl}_n$ or type $\mathfrak{o}_{2n+1}&#...

Useful Fields:

The article introduces a novel approach to deriving recurrence relations for off-shell Bethe vectors, a key aspect of quantum integrable models. This innovative method has the potential to significantly advance understanding in the field of integrable systems and may inspire further research on related quantum models and techniques. The work appears methodologically rigorous, building on established theoretical frameworks while contributing new insights. However, the immediate applicability to broader areas may be somewhat limited by its specialized focus.

100% Hallucination Elimination Using Acurai

The issue of hallucinations in large language models (LLMs) remains a critical barrier to the adoption of AI in enterprise and other high-stakes applications. Despite advancements in retrieval-augment...

Useful Fields:

The article presents a highly novel approach (Acurai) that claims to fully eliminate hallucinations in LLMs, a critical issue in AI applications. The methodological rigor appears strong, particularly the focus on reformatting input queries and context. By addressing a fundamental barrier to AI adoption, it may inspire significant advancements and attentiveness towards model reliability in future research endeavors. Overall, its relevance and potential impact within the field of AI are substantial, especially in high-stakes applications.

Interplay of intrinsic motion of partons and soft gluon emissions in Drell-Yan production studied with PYTHIA

Understanding the intrinsic transverse momentum (intrinsic- $k_T$ ) of partons within colliding hadrons, typically modeled with a Gaussian distribution characterized by a specific width (the int...

Useful Fields:

This article addresses a significant challenge in high-energy particle physics by exploring the non-perturbative aspects of parton dynamics and their impact on event generators like Pyhtia. The focus on intrinsic transverse momentum and the interplay of soft gluon emissions provides novel insights that could improve modeling accuracy in high-energy collisions. The methodological rigor in examining these complex interactions enhances its applicability and relevance.

Millisecond Pulsars in Globular Clusters and Implications for the Galactic Center Gamma-Ray Excess

We study the gamma-ray emission from millisecond pulsars within the Milky Way's globular cluster system in order to measure the luminosity function of this source population. We find that these pu...

Useful Fields:

The study provides valuable insights into the properties of millisecond pulsars and their role in explaining the Galactic Center Gamma-Ray Excess. Its methodological rigor in analyzing luminosity functions and the implications for gamma-ray emission are noteworthy. The novelty lies in addressing a significant astrophysical issue, contributing to our understanding of both pulsar populations and gamma-ray sources. However, while the findings are compelling, further investigation is needed to confirm or refute the proposed scenarios.

Structural characterization of the candidate Weyl semimetal CeGaGe

Weyl semimetals have a variety of intriguing physical properties, including topologically protected electronic states that coexist with conducting states. Possible exploitation of topologically protec...

Useful Fields:

The article presents a significant advancement in the characterization of a novel Weyl semimetal, CeGaGe, through meticulous experimental techniques such as single-crystal neutron diffraction. Its findings regarding the non-centrosymmetric structure are critical since they add to our understanding of both fundamental physics and potential technological applications of Weyl semimetals. The methodological rigor and the identification of emergent phenomena in non-centrosymmetric structures provide a strong foundation for future research in this area, making it a valuable contribution to the field.

Transformers Meet Relational Databases

Transformer models have continuously expanded into all machine learning domains convertible to the underlying sequence-to-sequence representation, including tabular data. However, while ubiquitous, th...

Useful Fields:

The article presents a novel approach that merges transformer models with relational databases, a significant development considering the increasing need for effective processing of structured data in machine learning. The methodological rigor is apparent in the comparison against various models and the focus on end-to-end learning, which addresses real-world challenges. The implications for both data processing and machine learning paradigms can inspire future research in related fields.

Stochastic Homogenisation of nonlinear minimum-cost flow problems

This paper deals with the large-scale behaviour of nonlinear minimum-cost flow problems on random graphs. In such problems, a random nonlinear cost functional is minimised among all flows (discrete ve...

Useful Fields:

The paper addresses a significant gap in the understanding of nonlinear minimum-cost flow problems within the stochastic framework. The use of Γ-convergence provides a novel approach to convergence results that are highly relevant for theoretical advancements in optimization and probabilistic modeling. Additionally, the examination of multi-species problems enhances its applicability across various domains. The methodological rigor demonstrated through the blow-up technique adds robustness to the findings, supporting its impact on future research developments.

ColonNet: A Hybrid Of DenseNet121 And U-NET Model For Detection And Segmentation Of GI Bleeding

This study presents an integrated deep learning model for automatic detection and classification of Gastrointestinal bleeding in the frames extracted from Wireless Capsule Endoscopy (WCE) videos. The ...

Useful Fields:

The article introduces a novel hybrid deep learning model that combines DenseNet121 and U-Net for a specific and clinically significant task—detecting and segmenting GI bleeding from Wireless Capsule Endoscopy videos. The fact that it achieved the highest performance in a competitive challenge demonstrates both its methodological rigor and its potential applicability in real-world diagnostics. Its accuracy of 80% is noteworthy, indicating a strong impact on clinical practice, while the release of the dataset enhances its contribution to future research, allowing other teams to build on this work.

Excitation spectrum of a double supersolid in a trapped dipolar Bose mixture

Dipolar Bose-Einstein condensates are excellent platforms for studying supersolidity, characterized by coexisting density modulation and superfluidity. The realization of dipolar mixtures opens intrig...

Useful Fields:

This article presents a novel investigation into the properties of dipolar Bose-Einstein condensates, specifically focusing on a double supersolid state. The exploration of the excitation spectrum in a trapped dipolar Bose mixture opens up new avenues for experimental research and deepens our understanding of superfluidity and supersolid phases. The rigorous analysis of various excitation modes adds to its methodological strength and provides practical insights relevant for future experiments.

AI's assigned gender affects human-AI cooperation

Cooperation between humans and machines is increasingly vital as artificial intelligence (AI) becomes more integrated into daily life. Research indicates that people are often less willing to cooperat...

Useful Fields:

This study introduces a novel angle to the examination of human-AI cooperation by exploring the impact of gender labels on the interactions, which has been previously unaddressed in the literature. The use of the Prisoner's Dilemma game provides a rigorous methodological framework, and the results underscore the persistent nature of human biases, suggesting implications for the design and regulation of AI technologies. The findings not only contribute to the understanding of human-AI dynamics but also provoke thoughts on societal norms and gender biases influencing such interactions.

Probing neutrino mass ordering with supernova neutrinos at NO $ν$ A including the effect of sterile neutrinos

In this work, we explore the possibility of probing the mass ordering sensitivity as a function of supernova distance in the context of the ongoing neutrino experiment NO $ν$ A. We provide a det...

Useful Fields:

The article presents a novel approach to investigating neutrino mass ordering using supernova neutrinos and addresses the impact of sterile neutrinos, which could lead to significant advancements in the understanding of neutrino properties. Its methodological rigor in presenting scenarios based on different mixing frameworks adds strength to the findings. However, the reliance on specific experimental setups may limit its broader applicability.