arXiv subjects

Manish Jain

Publications and source records attributed to Manish Jain.

At least 19 recordsLinked to original sources

arXiv2026

Comparative Assessment of Thermal Transport Theories: Dual-Channel Mechanism Dictates Heat Transport in Ultralow-$\kappa$ Materials

Anomalous heat transport in strongly anharmonic crystalline solids poses both a fundamental challenge to the theoretical understanding and an opportunity for thermoelectric and thermal barrier coating applications. Although Green-Kubo theory reproduces experimental thermal conductivity ($\kappa$) at high temperatures, it lacks microscopic insight and neglects the Bose-Einstein statistics of lattice vibrations. On the other hand, the conventional Boltzmann transport equation (BTE) framework, based on a phonon-gas picture, fails due to strong anharmonicity-induced overdamped phonons. Herein, the thermal transport properties in TlAgSe, a metal chalcogenide, and Cs$_2$PbI$_2$C$_2$, an all-inorganic layered Ruddlesden-Popper perovskite, are investigated by explicitly accounting for temperature-dependent lattice dynamics through machine learning interatomic potentials and employing the Wigner transport equation (WTE) framework. Crucially, heat conduction is governed not only by higher-order phonon scattering-dominated populations' transport channel described within the BTE, but also by a coherences' channel in the WTE framework arising from wave-like interbranch coherence between eigenstates. Incorporating four-phonon scattering, WTE predicts average room-temperature $\kappa$ values of 0.31 Wm$^{-1}$K$^{-1}$ (TlAgSe) and 0.38 Wm$^{-1}$K$^{-1}$ (Cs$_2$PbI$_2$C$_2$), in excellent agreement with experiments. Phonon scattering-rate analysis reveals strong coherences' contributions and prevalent overdamped phonon modes, demonstrating the breakdown of the conventional BTE framework based on the phonon quasiparticle picture with only first-order anharmonic perturbation. This computational approach provides a unified description of heat transport in ultralow-$\kappa$ materials, offering a basis for the rational design of phononic and thermoelectric devices.

cond-mat.mtrl-sci↗

arXiv2026

Barium Hexaferrite Thin Films as a Scalable Magnetic-Insulator Platform for Proximity-Engineered Spintronics

Rare-earth iron garnets, such as yttrium iron garnet (YIG) and thulium iron garnet (TmIG), are the benchmark magnetic insulators for spintronic and magnonic devices, but achieving usable perpendicular magnetic anisotropy (PMA) in these materials typically relies on substrate strain- engineering, requiring careful lattice-matching and specific growth conditions that constrain ma- terial accessibility. Here we establish sputter grown barium hexaferrite (BaFe12O19, BaM) as a magnetic-insulator alternative with strong intrinsic perpendicular anisotropy, requiring no strain engineering. X-ray diffraction, transmission electron microscopy and Raman spectroscopy confirm stoichiometric films with atomically smooth surfaces, while first-principles calculations corroborate a robust ferrimagnetic ground state. The films exhibit square out-of-plane hysteresis with a coercive field of nearly 0.1 T. Unlike rare-earth garnets, the perpendicular anisotropy in BaM is intrinsic to its magnetoplumbite crystal structure, arising independent of highly ordered strain. Interfaced with Pt and with exfoliated BiSbTeSe2 (BSTS), BaM induces proximity induced anomalous Hall trans- port, confirming efficient interfacial exchange coupling, while the BSTS/BaM heterostructure shows an additional Hall contribution suggestive of non-collinear interfacial spin textures. These results position BaM thin films as a scalable magnetic-insulator platform for spintronic and topological heterostructure devices beyond the constraints of garnet chemistry.

cond-mat.mtrl-sci↗

arXiv2026

Moir\'e Phonons and Emergent Exciton-Phonon Coupling in a Moir\'e Heterobilayer

Moir\'e superlattices have emerged as a new platform for engineering electronic and optical properties in van der Waals heterostructures, enabling control over correlated and excitonic phenomena. Yet the impact of moir\'e superlattices on exciton-phonon coupling remains largely unexplored. Here we demonstrate emergent, layer-selective coupling between moir\'e phonons and moir\'e excitons in angle-aligned WS2/WSe2 heterobilayers. Using a broadband terahertz phonon transducer, we coherently launch moir\'e phonons that resonantly perturb the excitonic states. We show that the exciton-phonon coupling is intrinsically modified by the moir\'e superlattice in a layer-selective manner. A driven oscillator model captures the dynamics, revealing three moir\'e phonon resonances with distinct coupling to the moir\'e excitons. First principles calculations show that many moir\'e phonon modes can arise with distinct strongly hybridized in-plane and out-of-plane vibrations in the moir\'e unit cells. The calculations further identify the three experimentally observed moir\'e phonons and their emergent characteristic coupling to the moir\'e excitons.

cond-mat.mtrl-sci↗

arXiv2026

Persistent singlet electronic character in the multiexcitonic triplet-pair state of strongly coupled pentacene singlet fission dimers

Singlet fission converts an optically excited singlet state into a spin-entangled triplet pair state (TT$_1$)$^1$ that can, in principle, yield two free triplets for photovoltaics and/or a polarized high spin state for quantum technologies. Synthetically tunable templates suggest that the above photophysics is governed by a subtle but poorly understood interplay of molecular motifs, geometry and structural fluctuations. Here, we investigate the (TT$_1$)$^1$ state in a library of conformationally flexible pentacenic dimers, where a (TT$_1$)$^1$-specific near-IR spectral feature is readily available. Using a suite of polarization-controlled impulsive optical spectroscopies, we find that (TT$_1$)$^1$ formation is specific to planar conformations and is accompanied by large nuclear reorganization in the (TT$_1$)$^1$ photoproduct. Introducing polarization anisotropy to track the electronic character of the (TT$_1$)$^1$ species, supported by screened configuration interaction based electronic structure theory, we find that significant singlet-triplet electronic mixing is persistent throughout its evolution. This behavior is universal across diverse bridging motifs and indicates that, once the triplet pair is strongly bound, neither substantial nuclear reorganization nor structural fluctuations on longer timescales are sufficient to suppress persistent singlet-triplet electronic mixing, such that triplet-pair decorrelation is outcompeted by its decay. Our observations establish polarization-selective pump-probe and anisotropy as a direct optical probe of triplet pair decorrelation, complementary to spin-selective measurements at longer timescales.

physics.chem-ph↗

arXiv2026

Layer-Polarization-Driven Metal-Insulator Transition in multi-band Graphene Moire' Superlattices

Graphene/hBN moir\'e superlattices provide a highly tunable platform for exploring emergent quantum phases in low-dimensional systems. Here, we investigate the moir\'e superlattice formed between hBN and ABA-stacked trilayer graphene (TLG), an inherently multi-band system. We demonstrate that the moir\'e potential is not merely a perturbation but a tool to hybridize the distinct massless and massive electronic sectors of TLG. By applying a perpendicular displacement field to tune layer polarization, we drive a fundamental reconstruction of the electronic band structure. Specifically, increasing the displacement field evolves the system from a multi-band regime to an effectively single-band regime at low energies, accompanied by a metal--insulator transition at the hole-doped secondary Dirac point. This transition originates from a redistribution of carriers across graphene layers that selectively enhances their coupling to the extrinsic moir\'e potential. Quantum capacitance measurements provide direct evidence for the suppression of the density of states at the hole-side secondary Dirac point, consistent with gap opening and the emergence of a displacement-field-tuned band gap. Theoretical calculations reproduce these observations and identify layer-selective coupling to the moir\'e potential as the underlying mechanism. These results demonstrate electrical control of an emergent insulating phase in a low-dimensional moir\'e system, and highlight that layer polarization and layer-selective coupling in multi-band moir\'e heterostructures provide a powerful route for engineering topological and correlated phases through band structure reconstruction and electron interactions.

cond-mat.mes-hall↗

arXiv2026

Real-World Challenges in Fake News Detection: Dealing with Posts by Cold Users

Social media serves as a primary source of information in the current digital era. Many people consume a vast range of information in a very short span, yet, amidst the stream of genuine information, fake news and rumors continue to spread. The need for effective detection models is becoming increasingly critical. Past user behavior and user engagement on a post are strong signals that SOTA approaches leverage for fake news detection and other post classification tasks. However, these approaches lean too heavily on knowing this past behavior, and thus suffer from a cold user problem, or users that are new or have minimal footprint on the platform. In this paper, we make three core contributions. We first establish the value of user behavior, both content and user-user interactions, in the task of fake news and rumor detection. We then establish the extensive prevalence of cold users in the real-world datasets, and show the need for newer algorithms considering cold users. We next propose a novel socially-aware context representation scheme - USER EVIDENCE NETWORK (UEN) - to detect the spread of misinformation and unverified information while efficiently navigating this cold user challenge. We introduce techniques that approximate missing or absent behavior data of a new user from existing users' interactions. By carefully addressing the cold user challenge, our work provides robust approaches targeting fake news and rumor detection for real-world platforms.

cs.SI↗

arXiv2026

Enhancing reasoning accuracy in large language models during inference time

Large Language Models (LLMs) often exhibit strong linguistic abilities while remaining unreliable on multi-step reasoning tasks, particularly when deployed without additional training or fine-tuning. In this work, we study inference-time techniques to improve the reasoning accuracy of LLMs. We systematically evaluate three classes of inference-time strategies: (i) self-consistency via stochastic decoding, where the model is sampled multiple times using controlled temperature and nucleus sampling and the most frequent final answer is selected; (ii) dual-model reasoning agreement, where outputs from two independent models are compared and only consistent reasoning traces are trusted; and (iii) self-reflection, where the model critiques and revises its own reasoning. Across all evaluated methods, we employ Chain-of-Thought (CoT) [1] prompting to elicit explicit intermediate reasoning steps before generating final answers. In this work, we provide a controlled comparative evaluation across three inference-time strategies under identical prompting and verification settings. Our experiments on LLM [2] show that self-consistency with nucleus sampling and controlled temperature value yields the substantial gains, achieving a 9% to 15% absolute improvement in accuracy over greedy single-pass decoding, well-suited for low-risk domains, offering meaningful gains with minimal overhead. The dual-model approach provides additional confirmation for model reasoning steps thus more appropriate for moderate-risk domains, where higher reliability justifies additional compute. Self-reflection offers only marginal improvements, suggesting limited effectiveness for smaller non-reasoning models at inference time.

cs.CL↗

arXiv2026

Signatures of moir\'e intralayer biexcitons and exciton-phason coupling in WSe2/WS2 heterostructures

Interactions among electronic and lattice degrees-of-freedom are foundational to various phases in condensed-matter physics, yet the dynamic interplay between excitonic and phononic quasiparticles represents an equivalent, underexplored frontier. Moir\'e superlattices provide an ideal platform for realizing these interactions by offering localized intralayer excitons (IALX) and ultralow-energy collective lattice modes, such as phasons. Here, by optically suppressing ultrafast charge-transfer (CT) to interlayer excitons in WSe2/WS2 heterostructures, we uncover dynamics of moir\'e IALX revealing long lifetimes ({\tau} > 1000 ps) arising from localized Wannier and in-plane CT nature. We then observe moir\'e intralayer intervalley biexcitons with binding energy ~ 16 meV, with long lifetimes due to moir\'e confinement. Furthermore, we find time-domain signatures of strong coupling between moir\'e-IALX and ~ 10 micro-eV phasons, evidenced as twist-angle-dependent GHz oscillations in IALX dynamics. Our findings establish moir\'e superlattices as interacting hybrid quantum systems and for engineering non-equilibrium phenomena, as well as for GHz-scale optoelectronics.

cond-mat.mes-hall↗

arXiv2025

Mortgage Language Model: Domain-Adaptive Pretraining with Residual Instruction, Alignment Tuning, and Task-Specific Routing

Large Language Models (LLMs) demonstrate exceptional capabilities across general domains, yet their application to specialized sectors such as mortgage finance requires domain-specific knowledge augmentation while preserving instruction-following fidelity. We present MortgageLLM, a novel domain-specific large language model that addresses this dual challenge. It is developed using a dual-track specialization framework from a single base model (LLaMA-3.1-8B). We opted for this dual-expert approach as a single multi-task model suffers from performance trade-offs, where optimizing for structured tasks (via SFT) degrades conversational fidelity (via DPO). Our dual-track method solves this by creating two specialists, allowing each to be optimally trained for its distinct capability. Our approach applies the instruction residual technique to restore instruction-following capabilities post-domain adaptation without supervised fine-tuning. We contribute: (1) application of this residual technique to the highly specialized mortgage finance domain; (2) a dual-expert architecture combining a conversational Q&A model and a structured task model for classification and summarization; and (3) an intelligent task routing mechanism using few-shot classification performed by one of the expert models itself. We validate our approach on domain-specific benchmarks, where our final model (MLM v2) significantly outperforms the base LLaMA-3.1-8B-Instruct, achieving an LLM-as-a-Judge summarization score of 4.58 (vs. 3.99), a Q&A score of 4.09 (vs. 4.0), and a classification score of 2.6 (vs. 1.2). On semantic similarity, our model achieved a BERTScore of 0.77 for summarization (vs. 0.74), 0.68 for Q&A (vs. 0.58), and 0.75 for classification (vs. 0.73), substantially outperforming baseline approaches.

cs.CL↗

arXiv2025

Reasoning-Guided Claim Normalization for Noisy Multilingual Social Media Posts

We address claim normalization for multilingual misinformation detection - transforming noisy social media posts into clear, verifiable statements across 20 languages. The key contribution demonstrates how systematic decomposition of posts using Who, What, Where, When, Why and How questions enables robust cross-lingual transfer despite training exclusively on English data. Our methodology incorporates finetuning Qwen3-14B using LoRA with the provided dataset after intra-post deduplication, token-level recall filtering for semantic alignment and retrieval-augmented few-shot learning with contextual examples during inference. Our system achieves METEOR scores ranging from 41.16 (English) to 15.21 (Marathi), securing third rank on the English leaderboard and fourth rank for Dutch and Punjabi. The approach shows 41.3% relative improvement in METEOR over baseline configurations and substantial gains over existing methods. Results demonstrate effective cross-lingual generalization for Romance and Germanic languages while maintaining semantic coherence across diverse linguistic structures.

cs.CL↗

arXiv2025

Emergent Rashba spin-orbit coupling in bulk gold with buried network of nanoscale interfaces

The Rashba effect, which plays a crucial role in fundamental materials physics and potential spintronics applications, has been engineered in diverse systems, including semiconductor quantum wells, oxide heterostructures, metallic surfaces, topological insulators, ferroelectrics, etc. However, generating it in systems that preserve bulk inversion symmetry (BIS), for example, in bulk metals, has not been possible so far. We demonstrate a unique strategy to introduce and tune Rashba spin-orbit interaction (SOI) to unprecedented magnitudes in inversion-symmetric solids, by incorporating ultra-small silver nanoparticles in bulk gold. The near-identical lattice constants of Ag and Au allowed dense packing of the Ag/Au hetero-interfaces without compromising the global BIS. By varying the density of embedded nanoparticles, we generate Rashba SOI in a bulk metal with a coupling strength of ~15 meV.Angstrom, higher than any known system preserving BIS globally, and up to ~20 times increase in the spin-orbit scattering rate. We argue that the combined effect of charge-transfer at the interfaces and polaronic localization enhances the SOI.

cond-mat.mes-hall↗

arXiv2025

Double excitations in molecules

Double excitations in organic molecules have garnered significant interest as a result of their importance in singlet fission and photophysics. These excitations play a crucial role in understanding the photoexcitation processes in polyenes. To describe photoexcited states with both single and double excitation character, we use a first-principles many-body theory that combines the GW / Bethe-Salpeter equation and the configuration interaction (CI) methods. Specifically, we develop and employ two CI-based methods: screened configuration interaction singles and doubles (scrCISD) and screened configuration interaction singles with perturbative doubles (scrCIS(D)), applied to an effective many-body Hamiltonian that incorporates screening. We apply these methods to Thiel's set of molecules, which exhibit excited states predominantly characterized by single excitations with a partial double excitation character. Our results indicate that the scrCISD method systematically underestimates the excitation energies compared to the best theoretical estimates, while the scrCIS(D) method shows good agreement with these estimates. Furthermore, we used the scrCISD method to calculate the binding energies of the dominantly doubly excited correlated triplet pair states, $\mathrm{TT^1}$, in pentacene dimers, finding that the $\mathrm{TT^1}$ binding energies agree well with empirical calculations.

physics.chem-ph↗

arXiv2025

Symmetries in zero and finite center-of-mass momenta excitons

We present a symmetry-based framework for the analysis of excitonic states, incorporating both time-reversal and space-group symmetries. We demonstrate the use of time-reversal and space-group symmetries to obtain exciton eigenstates at symmetry-related center-of-mass momenta in the entire Brillouin zone from eigenstates calculated for center-of-mass momenta in the irreducible Brillouin zone. Furthermore, by explicitly calculating the irreducible representations of the little groups, we classify excitons according to their symmetry properties across the Brillouin zone. Using projection operators, we construct symmetry-adapted linear combinations of electron-hole product states, which block diagonalize the Bethe-Salpeter equation (BSE) Hamiltonian at both zero and finite exciton center-of-mass momenta. This enables a transparent organization of excitonic states and provides direct access to their degeneracies, selection rules, and symmetry-protected features. As a demonstration, we apply this formalism to monolayer MoS$_2$, where the classification of excitonic irreducible representations and the block structure of the BSE Hamiltonian show excellent agreement with compatibility relations derived from group theory. Beyond this material-specific example, the framework offers a general and conceptually rigorous approach to the symmetry classification of excitons, enabling significant reductions in computational cost for optical spectra, exciton-phonon interactions, and excitonic band structure calculations across a wide range of materials.

cond-mat.mtrl-sci↗

arXiv2025

Enhanced Phonon-Assisted Tunneling in Metal -- Twisted Bilayer Graphene Junctions

We report planar tunneling spectroscopy measurements on metal-WSe$_2$-twisted bilayer graphene heterostructures across a broad range of gate and bias voltages. The observed experimental features are attributed to phonon-assisted tunneling and the significantly high density of states within the moir\'e bands. A notable finding is the enhanced phonon-assisted tunneling in twisted bilayer graphene compared to Bernal bilayer graphene, which arises from a more relaxed in-plane momentum matching criterion. Theoretical calculations of phonon dispersions enable us to identify low-energy phonon modes in both Bernal and twisted bilayers of graphene, thereby elucidating the underlying mechanism of tunneling. Our results establish planar tunneling as a versatile tool to further understand electron-phonon coupling in twisted van der Waals materials.

cond-mat.mes-hall↗

arXiv2025

Probing Phonon Modes in Reconstructed twisted Homo and Hetero Bilayer System

Twist angle engineering in van der Waals homo and hetero-bilayers introduces profound modifications in their electronic, optical and mechanical properties due to lattice reconstruction. In these systems, the interlayer coupling and atomic rearrangement strongly depend on the twist angle, leading to the formation of periodic Moire superlattices. At small twist angles, significant lattice relaxation results in the emergence of domain structures separated by one dimensional soliton networks, influencing electronic band structures and phonon modes. Here we systematically investigate the impact of lattice reconstruction on phonon renormalization in twisted bilayer graphene (TBLG,homo) and graphene-hBN Moire superlattices(hetero). Using Raman spectroscopy, we identify distinct phonon behaviours across different twist angle regimes. In TBLG, we observe the evolution of the G peak, including broadening, splitting, and the emergence of additional peaks in the small angle range 0.3 to 1 degree, attributed to Moire modified phonon interactions. At large twist angles, the peaks gradually merge back into a single feature, reflecting the reduced impact of lattice reconstruction. Similarly, in hBN graphene Moire superlattices, we detect Moire induced Raman peaks above and below the G peak, while the central G peak remains largely invariant to twist angle variation. The theoretical calculations uncover Moire phonon modes originating from different stacking regions providing insights into phonon renormalization. Our results establish a direct link between twist angle, lattice reconstruction, Moire phonons, and interlayer coupling, offering a fundamental framework for understanding phonon engineering in twisted bilayer systems. These findings pave the way for controlling phononic, optoelectronic and heat flow properties in next generation van der Waals heterostructures.

cond-mat.mes-hall↗

arXiv2025

Topo Goes Political: TDA-Based Controversy Detection in Imbalanced Reddit Political Data

The detection of controversial content in political discussions on the Internet is a critical challenge in maintaining healthy digital discourse. Unlike much of the existing literature that relies on synthetically balanced data, our work preserves the natural distribution of controversial and non-controversial posts. This real-world imbalance highlights a core challenge that needs to be addressed for practical deployment. Our study re-evaluates well-established methods for detecting controversial content. We curate our own dataset focusing on the Indian political context that preserves the natural distribution of controversial content, with only 12.9% of the posts in our dataset being controversial. This disparity reflects the true imbalance in real-world political discussions and highlights a critical limitation in the existing evaluation methods. Benchmarking on datasets that model data imbalance is vital for ensuring real-world applicability. Thus, in this work, (i) we release our dataset, with an emphasis on class imbalance, that focuses on the Indian political context, (ii) we evaluate existing methods from this domain on this dataset and demonstrate their limitations in the imbalanced setting, (iii) we introduce an intuitive metric to measure a model's robustness to class imbalance, (iv) we also incorporate ideas from the domain of Topological Data Analysis, specifically Persistent Homology, to curate features that provide richer representations of the data. Furthermore, we benchmark models trained with topological features against established baselines.

cs.SI↗

arXiv2025

Even-denominator fractional quantum Hall states in the zeroth Landau level of ABA trilayer graphene

Even-denominator fractional quantum Hall states (FQHSs) at half filling are of particular interest because they can host non-Abelian quasiparticles. Here we report the emergence of such states in the zeroth Landau level ($N=0$) of ABA trilayer graphene (TLG), challenging the conventional expectation that they are confined to the first excited Landau level. We observe robust incompressible states at $\nu=7/2$, $9/2$, and $5/2$ with their associated Levin--Halperin daughter states: $\nu=59/17$ and $46/13$ near $7/2$; $\nu=58/13$ and $77/17$ near $9/2$; and $\nu=43/17$ near $5/2$. These states appear exclusively within a finite displacement-field window coincident with crossings between symmetry-broken $N=0$ Landau levels carrying distinct isospin indices. The quantitative correspondence between the calculated crossing loci and the experimentally determined stability regions identifies Landau-level mixing as the microscopic origin. We attribute the stabilization of these even-denominator states to inversion-symmetry breaking in TLG, which enhances valley-resolved Landau-level hybridization and renormalizes short-range Coulomb interactions. Our results expand the landscape of even-denominator FQHSs to multilayer graphene and establish TLG as a tunable platform for realizing non-Abelian anyons.

cond-mat.mes-hall↗

arXiv2024

Controlling particle-hole symmetry of fractional quantum hall states in trilayer graphene

We present a detailed experimental study of the particle-hole symmetry (PHS) of the fractional quantum Hall (FQH) states about half filling in a multiband system. Specifically, we focus on the lowest Landau level of the monolayer-like band of Bernal stacked trilayer graphene (TLG). In pristine TLG, the excitation energy gaps, Land\'e g-factor, effective mass, and disorder broadening of the odd-denominator FQH states are identical to their hole-conjugate counterpart. This precise PH symmetry stems from the lattice mirror symmetry that precludes Landau-level mixing. Introducing a non-zero displacement field $D$ disrupts this mirror symmetry, facilitating the hybridization between the monolayer-like and bilayer-like Landau levels. This inter-band coupling enhances the Landau level mixing factor $\eta$ and activates three-body interactions -- both of which explicitly break the PHS of FQHs. As a result, conventional FQHs are completely destabilized, offering a route to engineer symmetry breaking of FQHs in a controlled way. We establish that the PHS breaking in TLG is of extrinsic origin and is fundamentally distinct from the intrinsic, interaction-driven symmetry breaking observed in the lowest Landau levels of single-layer and bilayer graphene.

cond-mat.mes-hall↗