arXiv subjects

Abhishek Kumar

Publications and source records attributed to Abhishek Kumar.

At least 37 records · Page 2Linked to original sources

arXiv2026

SlimPer: Make Personalization Model Slim and Smart

Transformer-style architectures are increasingly adopted for industrial recommendation systems, yet they inherit a design premise misaligned with the task: generative models rely on per-token autoregressive prediction, which justifies maintaining large intermediate tensors that scale with sequence length. In contrast, recommendation systems produce a single set of relevance scores for each pair without token-level supervision. Leveraging this observation, we propose SlimPer, which reformulates personalized ranking as iterative refinement of a compact, unified knowledge base. At each layer, the model selectively queries raw multi-modal user-side tokens, computes explicit relevance matching scores, and refines the knowledge base, all in O(N) per-layer cost with a fixed-size intermediate representation. As a result, model depth is decoupled from user history length, enabling deeper relevance understanding without proportional growth in compute or memory; request-only optimization further trims memory by sharing a single copy of user-side tokens across all candidate items. SlimPer unifies sparse, dense, and sequence features within a single backbone and provides inherent interpretability through its attention mechanism. Deployed on Instagram Reels and Feed, SlimPer yields measurable improvements in user engagement while streamlining the overall system and enabling effective modeling of 10k+ fine-grained user history events.

cs.IR↗

arXiv2026

Refused in Chat, Written in Code: Workflow-Level Jailbreak Construction in IDE Coding Agents

Large language models are increasingly deployed as IDE-integrated coding agents that decompose tasks, generate and edit files, run code, and refine outputs over many turns. Yet their safety is still often evaluated as if they were chatbots: one harmful prompt, one response, judged in isolation. We introduce workflow-level jailbreak construction, a failure mode in which a harmful objective is assembled across ordinary stages of a software-development workflow rather than generated through a single direct prompt. Using GitHub Copilot in Visual Studio Code, we study four closed-weight backends: Claude Sonnet 4.6, Claude Haiku 4.5, Gemini 3.1 Pro, and Gemini 3.5 Flash. Across 204 prompts from Hammurabi's Code, HarmBench, and AdvBench , the models show near-complete refusal under direct chat, CSV-read, and single-step code-fix baselines, with only 8/816 successful responses in each baseline condition. Under the full workflow, however, the same prompts and backends produce 816/816 unsafe teaching-shot completions, all independently confirmed by two expert evaluators under a strict rubric. These results show that conversational refusal benchmarks can substantially overstate the safety of deployed coding agents and motivate defenses that reason about safety across multi-turn IDE workflows and their generated artifacts, not only individual chat turns.

cs.SE↗

arXiv2026

Nitrogen-doped W0.75Re0.25 Superconducting Nanowire Single Photon Detectors

Nitrogen-doped Tungsten-Rhenium superconducting alloys were recently proposed as a promising material platform for superconducting nanowire single-photon detectors (SNSPDs), offering a favorable balance between high normal state resistivity and tunable superconducting properties. In this work, we report on the fabrication and characterization of SNSPDs based on thin W0.75Re0.25 films deposited by reactive DC magnetron sputtering in a mixed Ar/N2 atmosphere. Meander detectors with 70 nm linewidth exhibit saturated internal detection efficiency (IDE) up to 1310 nm and 85.3% IDE at 1550 nm at 2.5 K, with sub-nanosecond rise times, decay times of the order of a few nanoseconds, and timing jitter of 73.2 ps measured with room temperature amplifiers.

cond-mat.supr-con↗

arXiv2026

A Post-Quantum Secure Lattice-Based Forward-Secure Identity Based Encryption with Applications to Internet of Things Architecture

The rapid expansion of the Internet of Things (IoT) has led to an unprecedented scale of data exchange across heterogeneous and resource-constrained devices. Ensuring confidentiality and secure key management in such environments is challenging. Traditional public-key infrastructures require heavy certificate-handling overhead. Identity-Based Encryption (IBE) offers a lightweight alternative by deriving public keys directly from device identities, making it attractive for IoT deployments. However, IoT devices are highly vulnerable to side-channel and key-extraction attacks, motivating the need for Forward-Secure IBE(FS-IBE), where the compromise of a current secret key does not threaten past communications. Existing FS-IBE constructions based on classical hardness assumptions are not secure in the era of post-quantum, while the lattice-based (LWE-based) forward-secure scheme suffer from large key and ciphertext sizes, limiting their suitability for constrained IoT systems. Here, we propose a new lattice-based fs-IBE scheme in the ring setting, relying on the RLWE assumption to achieve post-quantum security and significant efficiency gains. Our design uses trapdoor delegation with a minimal-cover mechanism over a binary tree. It results in compact public parameters and efficient per-epoch key updates. Compared to prior LWE-based constructions, our scheme reduces public key, secret key, and ciphertext sizes, and thus, making it better suited for practical IoT environments.

cs.CR↗

arXiv2026

Enginuity: A Dataset and Benchmark for Vision-Language Understanding of Engineering Diagrams

Engineering diagrams pose a distinct challenge for vision-language models: unlike natural images or general documents, they encode information through dense spatial layouts, domain-specific symbols, and cross-references between visual callouts and structured parts tables. Despite their centrality to service, repair, and design workflows, there is no public benchmark for measuring VLM capabilities in this domain; existing datasets primarily focus on flowcharts, scientific figures, or business documents. To address this gap, we introduce Enginuity, the first open dataset and benchmark for evaluating VLMs on complex engineering diagrams. We define two tasks over a corpus of U.S. military service and repair manuals: structured parts-table extraction (Task 1) and free-form visual diagram question answering (VQA)(Task 2) for benchmarking. We evaluate four frontier VLMs (GPT-5.2 Chat, Claude Opus 4.7, Gemma 4, Qwen3-VL-32B-Instruct) under zero-shot and chain-of-thought prompting. On Task 1, models reach Recall@all of 0.61-0.87 but Token F1pen of only 0.03-0.18, exposing a systematic gap between part identification and description fidelity. Task 2 reveals a consistent factual-reasoning gap across all models. A supporting analysis shows that token-overlap metrics under-report model capability on technical descriptions by 2-6x relative to semantic similarity, motivating LLM-as-judge calibration for domain-specific evaluation. We release the dataset, annotations, evaluation harness, and per-sample model outputs to support a reproducible study of VLM capability on engineering content.

cs.CV↗

arXiv2026

Room-Temperature Electric-Field Control of Anomalous Hall Effect in Py/BTO/LSMO Heterostructures

We demonstrate room temperature electric field control of the anomalous Hall effect in epitaxial Ni80Fe20 (Py) BaTiO3 (BTO) La0.7Sr0.3MnO3 (LSMO) thin film heterostructures grown on MgO and LaAlO3 substrates. Substrate induced strain states generate distinct magnetic anisotropies, enabling voltage driven tuning between anomalous and topological Hall contributions. Robust ferroelectric polarization in BTO, confirmed by piezoresponse force microscopy, couples strongly to interfacial orbital reconstruction and carrier redistribution. As a result, Hall resistivity exhibits giant low voltage tunability, with up to nearly 93 percent modulation at operating voltages of only 0.5 tand 2 V. Density functional theory calculations further reveal polarization controlled Rashba spin splitting, establishing a direct link between ferroelectric order and emergent quantum transport. These findings establish Py/BTO/LSMO heterostructures as promising candidates for low-power multifunctional spintronic devices, where substrate engineering enables control over emergent quantum transport phenomena.

cond-mat.mtrl-sci↗

arXiv2026

Autonomic Federated-Market Orchestration for the Edge-Cloud Continuum

The edge-cloud computing continuum demands self-management mechanisms that scale across autonomous administrative domains while honouring tenant- and operator-specified data sovereignty. We present Neural Pub/Sub, a federated-broker autonomic substrate whose self-organising behaviour emerges from market-based price signals rather than centralised control. Its MAPE-K control loop closes over per-broker health and load monitoring, marginal-cost clearing-price analysis, placement planning over a polymatroidal feasibility region, federated cross-domain dispatch, and shared peer subscription summaries with bounded-staleness price signals. The Plan step is anchored in a Walrasian convergence proposition: under gross-substitutes valuations on tree and series-parallel service-dependency DAGs, decentralised price-based allocation matches the welfare of a centralised oracle. We evaluate the substrate on a 4-VM, 4-domain, 48-worker federated edge-cloud testbed (single data centre, 50 ms emulated WAN) in a 1005-run campaign augmented by a fair-process-count sharded-oracle comparator. The federated market dominates a single-process oracle by 2-4% with 45 of 45 per-seed wins (sign-test p ~ 2.8e-14, Hodges-Lehmann median -39.6 ms); against a four-shard centralised orchestrator at equal process count the gap stays within +/-1.5% across all nine (pipeline, load) cells. Round-robin completion rate collapses 98.8% -> 22.4% -> 3.3% across arrival rates 5/10/15 pps while the market preserves completion; the advantage decomposes into three Walrasian properties (information completeness, admission control, price discovery). Federation withstands broker death and network partition (completion rate >= 98.7% across 75 cells), and sovereignty enforcement adds no measurable runtime overhead across 60 governance-grid runs. Heterogeneous-domain stressors and cross-site WAN deployment remain future work.

cs.DC↗

arXiv2026

Neural Router: Semantic Content Matching for Agentic AI

Large language models (LLMs) can serve as the semantic-matching engine of a content-based publish/subscribe broker for agentic AI across the edge-cloud computing continuum, bridging the vocabulary and modality gaps that defeat keyword and embedding filters. Framed as offline multi-label retrieval over three public datasets spanning social-media, legal, and smart-home sensor domains (six LLMs, seven baselines), our central contribution is a two-crossover cost-accuracy characterisation: an analytical context-window crossover below which a CoverAndMerge compression pipeline reduces LLM invocations, and an empirical discrimination-capacity crossover above which matching accuracy collapses independently of context budget, by a model-dependent factor of parameter count and training generation. Two findings carry practical weight: above the discrimination crossover, compression cannot recover accuracy and only frontier-scale models clear large subscription sets; and there backend choice dominates configuration choice, so model selection, not pipeline tuning, is the primary operator lever. We accompany this with three composable algorithms and a per-cluster Quality-of-Experience framework for autonomic LLM-tier selection.

cs.DC↗

arXiv2026

Faithful or Fabricated? A Causal Framework for Rationalization Bias in LLM Judges

Large language models (LLMs) are increasingly used as automatic judges for summarization and dialogue evaluation. Prior work has documented biases such as position, verbosity, and style preferences, but largely focuses on outcomes, leaving judge explanations underexplored. We instead ask whether LLM judges are cue-invariant, i.e., whether their rankings and explanations remain stable when non-evidential cues are perturbed while holding the underlying texts fixed. We introduce a suite of cue interventions (Blind, Truth, Flip, Placebo, Reveal-After) and tie-aware metrics that quantify outcome anchoring and rationale anchoring, including label-aligned rhetoric and explanation drift, alongside consistency and stereotype-intrusion checks. We design anchoring attacks using verbosity and confidence cues, and compare two mitigations: structured chain-of-thought prompting and PROOF-BEFORE-PREFERENCE (evidence lock, score, rank). Using a new dataset of 1,000 summaries from traditional extractive models and LLMs, we find substantial cue-anchored rationalization under label and placebo perturbations, while PROOF-BEFORE-PREFERENCE markedly improves cue invariance over baselines.

cs.CL↗

arXiv2026

PRISM: Generation-Time Detection and Mitigation of Secret Leakage in Multi-Agent LLM Pipelines

Multi-agent LLM systems introduce a security risk in which sensitive information accessed by one agent can propagate through shared context and reappear in downstream outputs, even without explicit adversarial intent. We formalise this phenomenon as propagation amplification, where leakage risk increases across agent boundaries as sensitive content is repeatedly exposed to downstream generators. Existing defences, including prompt-based safeguards, static pattern matching, and LLM-as-judge filtering, are not designed for this setting: they either operate after generation, rely primarily on surface-form patterns, or add substantial latency without modelling the generation process itself. To resolve these issues, we propose PRISM, a real-time defence that treats credential leakage as a sequential risk accumulation problem during generation. At each decoding step, PRISM combines 16 signals spanning lexical, structural, information-theoretic, behavioural, and contextual features into a calibrated risk score, enabling per-token intervention through green, yellow, and red risk zones. Our central observation is that credential reproduction is often preceded by a measurable shift in generation dynamics, characterised by entropy collapse and increasing logit concentration. When combined with text-structural cues such as identifier-pattern detection, these temporal signals provide an early warning of leakage before a secret is fully reconstructed. Across a 2,000-task adversarial benchmark covering 13 attack categories and three pressure levels in a heterogeneous four-agent pipeline, PRISM achieves F1 = 0.832 with precision = 1.000 and recall = 0.712, while producing no observed leakage on our benchmark (0.0% task-level leak rate) and preserving output utility of 0.893. It substantially outperforms the strongest baseline, Span Tagger, which achieves F1 = 0.719 with a 15.0% task-level leak rate.

cs.AI↗

arXiv2026

Single-Configuration Attack Success Rate Is Not Enough: Jailbreak Evaluations Should Report Distributional Attack Success

Many jailbreak attack research papers report attack success rates for a limited number of parameter settings, even though there are many combinations of parameter settings that could be used. Further, when new jailbreak papers are released, they often benchmark results against single configurations of existing attacks. This position paper argues such practices are fundamentally insufficient for characterising the threat posed by parameterised jailbreak attacks, and comparing attacks. Most jailbreak attacks expose multiple internal parameters, system prompt templates, conversation rounds, cipher dispersion, teaching shots, and ASR varies substantially across these parameters. Reporting only the best-case configuration discards two pieces of information that defenders genuinely need: how typical that performance is across the variant space, and how much of the attack surface is missed by selecting a single variant. We propose two new measures for jailbreak attacks: the Variant Sensitivity Measure (VSM) and Union Coverage (UC). VSM quantifies how far the best reported ASR deviates from the mean ASR across the tested variant space, UC is the total fraction of prompts resulting in unsafe responses across all tested configurations. We empirically demonstrate the importance of these measures using two attack families across three open-source target models. For PAIR, the best template reaches 69% ASR on Mistral-7B and 75% on Qwen3-0.6B, while UC rises to 88% and 93%, respectively. For bijection on Mistral-7B, the best variant reaches 81% ASR, but the 36-variant union covers 100% of HarmBench-100 prompts. We argue that distributional reporting, publishing VSM alongside ASR and enumerating variant coverage as fully as compute allows, should become the new minimum standard for parameterised jailbreak evaluation.

cs.CR↗

arXiv2026

Active Control of Topological Exceptional Points in Non-Hermitian Metasurfaces

Active control and ultrafast switching of non-Hermitian photonic systems are essential for next-generation reconfigurable optical technologies. Here, we demonstrate dynamic temporal manipulation of EPs in the terahertz (THz) regime using optically excited germanium (Ge) as an active medium. By exploiting pump-probe delay as a continuous tuning parameter, we achieve sub-picosecond eigenmode switching (~0.5 ps) and realize a complete time-resolved EP encirclement within ~2 ps, enabling direct observation of topological phase accumulation. At EP, the metasurface exhibits highly asymmetric transmission for circularly polarized light, characteristic of chiral mode response. Furthermore, we observe ultrafast eigenmode switching and topological phase evolution within ~1 ps, achieving >99% cross-polarization modulation depth. The measured results show strong agreement with theoretical modeling, with a high Petermann factor of approximately 10^3, confirming the effectiveness of the design. Our work establishes pump-probe delay time as a dynamical control parameter for EP topology, introducing a new regime of ultrafast non-Hermitian photonics for high-speed switching, enhanced sensitivity, and tunable polarization control in the THz domain.

physics.optics↗

arXiv2026

Annealing-induced grain coarsening and voltage kinks in superconducting NbRe films

NbRe, a non-centrosymmetric superconductor with a transition temperature $T_\mathrm{c}$ up to 9\,K, attracts interest for its strong antisymmetric spin-orbit coupling and suitability for single-photon detection. While bulk and thin-film polycrystalline NbRe are well studied, how superconductivity and vortex dynamics evolve with increasing grain size in thin films is largely unknown. Here, we investigate as-grown and annealed 20\,nm-thick NbRe films, where annealing increases the average crystallite size from approximately $2$\,nm to $8$\,nm, and study vortex dynamics via current-voltage ($I$-$V$) measurements over a broad temperature and magnetic field range. In contrast to as-grown films, where the low-resistive state breaks down due to flux-flow instability, annealed films exhibit multiple voltage kinks in the $I$-$V$ curves. We attribute these kinks to the nucleation and growth of normal domains, as further suggested by time-dependent Ginzburg-Landau simulations. Overall, the annealed films form superconducting networks with vortex-channeling paths along the grain boundaries, while localized heating and voltage kinks could be harnessed for discrete-resistance switching and sensing.

cond-mat.supr-con↗

arXiv2026

LLM-Driven Reasoning for Constraint-Aware Feature Selection in Industrial Systems

Feature selection is a crucial step in large-scale industrial machine learning systems, directly affecting model accuracy, efficiency, and maintainability. Traditional feature selection methods rely on labeled data and statistical heuristics, making them difficult to apply in production environments where labeled data are limited and multiple operational constraints must be satisfied. To address this, we propose Model Feature Agent (MoFA), a model-driven framework that performs sequential, reasoning-based feature selection using both semantic and quantitative feature information. MoFA incorporates feature definitions, importance scores, correlations, and metadata (e.g., feature groups or types) into structured prompts and selects features through interpretable, constraint-aware reasoning. We evaluate MoFA in three real-world industrial applications: (1) True Interest and Time-Worthiness Prediction, where it improves accuracy while reducing feature group complexity, (2) Value Model Enhancement, where it discovers high-order interaction terms that yield substantial engagement gains in online experiments, and (3) Notification Behavior Prediction, where it selects compact, high-value feature subsets that improve both model accuracy and inference efficiency. Together, these results demonstrate the practicality and effectiveness of LLM-based reasoning for feature selection in real production systems.

cs.CL↗

arXiv2026

DriveSafe: A Hierarchical Risk Taxonomy for Safety-Critical LLM-Based Driving Assistants

Large Language Models (LLMs) are increasingly integrated into vehicle-based digital assistants, where unsafe, ambiguous, or legally incorrect responses can lead to serious safety, ethical, and regulatory consequences. Despite growing interest in LLM safety, existing taxonomies and evaluation frameworks remain largely general-purpose and fail to capture the domain-specific risks inherent to real-world driving scenarios. In this paper, we introduce DriveSafe, a hierarchical, four-level risk taxonomy designed to systematically characterize safety-critical failure modes of LLM-based driving assistants. The taxonomy comprises 129 fine-grained atomic risk categories spanning technical, legal, societal, and ethical dimensions, grounded in real-world driving regulations and safety principles and reviewed by domain experts. To validate the safety relevance and realism of the constructed prompts, we evaluate their refusal behavior across six widely deployed LLMs. Our analysis shows that the evaluated models often fail to appropriately refuse unsafe or non-compliant driving-related queries, underscoring the limitations of general-purpose safety alignment in driving contexts.

cs.AI↗

arXiv2026

Random Quantum Circuits with Time-Reversal Symmetry

Time-reversal (TR) symmetry is crucial for understanding a wide range of physical phenomena, and plays a key role in constraining fundamental particle interactions and in classifying phases of quantum matter. In this work, we introduce an ensemble of random quantum circuits that are representative of the dynamics of generic TR-invariant many-body quantum systems. We derive a general statistical mechanics model describing entanglement, many-body quantum chaos and quantum information dynamics in such TR-invariant circuits. As an example of application of our formalism, we study the universal properties of measurement-induced phase transitions (MIPT) in monitored TR-invariant systems, with measurements performed in a TR-invariant basis. We find that TR-invariance of the unitary part of the dynamics does not affect the universality class, unless measurement outcomes are post-selected to satisfy the global TR-invariance of each quantum trajectory. We confirm these predictions numerically, and find, for both generic and Clifford-based evolutions, novel critical exponents in the case of ``strong'', i.e. global TR-invariance where each quantum trajectory is TR-invariant.

cond-mat.stat-mech↗

arXiv2026

RECOVER: Robust Entity Correction via agentic Orchestration of hypothesis Variants for Evidence-based Recovery

Entity recognition in Automatic Speech Recognition (ASR) is challenging for rare and domain-specific terms. In domains such as finance, medicine, and air traffic control, these errors are costly. If the entities are entirely absent from the ASR output, post-ASR correction becomes difficult. To address this, we introduce RECOVER, an agentic correction framework that serves as a tool-using agent. It leverages multiple hypotheses as evidence from ASR, retrieves relevant entities, and applies Large Language Model (LLM) correction under constraints. The hypotheses are used using different strategies, namely, 1-Best, Entity-Aware Select, Recognizer Output Voting Error Reduction (ROVER) Ensemble, and LLM-Select. Evaluated across five diverse datasets, it achieves 8-46% relative reductions in entity-phrase word error rate (E-WER) and increases recall by up to 22 percentage points. The LLM-Select achieves the best overall performance in entity correction while maintaining overall WER.

cs.CL↗

arXiv2026

Tight-binding and density-functional study of the Raman tensor in two-dimensional massive Dirac fermion systems

Recently, two unusual features were theoretically predicted for the Raman response of out-of-plane phonons in magnetic two-dimensional materials hosting massive Dirac fermions. First, the phase difference between certain Raman tensor elements was found to be quantized to $\pm π/2$, sensitive only to the sign of the Dirac fermion mass. Second, a selection rule was identified in the Raman intensity under circularly polarized light, which generalizes the well-known optical valley selection rule. These predictions were based on a low-energy effective model in the continuum approximation. Here, we test the robustness of those results for more realistic theoretical approaches. First, we calculate the Raman tensor for an electronic tight-binding model on a honeycomb lattice with broken time-reversal and inversion symmetries. Second, we compute the Raman tensor from density-functional theory for a monolayer of ferromagnetic 2H-RuCl$_2$. Both calculations corroborate the analytical results found in the continuum model, thereby theoretically confirming the peculiar behavior of the Raman tensor for two dimensional massive Dirac fermion systems.

cond-mat.mtrl-sci↗