arXiv subjects

Nikolay Ivanov

Publications and source records attributed to Nikolay Ivanov.

At least 19 recordsLinked to original sources

arXiv2026

Enoki: Efficient Multi-Level Hallucination Detection

Ensuring factuality remains a critical challenge for deploying LLMs in high-stakes settings. Existing hallucination detectors usually operate at a single level: claim-level methods provide interpretable factual units, while span-level methods localize unsupported text. Bridging these views is costly, as LLM-heavy pipelines require multiple decomposition and verification calls, and modular systems need additional claim-to-span alignment. We propose Enoki, an Open Information Extraction framework for multi-level hallucination detection. Enoki extracts text-anchored relational facts, verifies them against evidence, and projects unsupported facts back to hallucinated spans. This shared representation enables claim-level verification and span-level localization without requiring separate alignment. Enoki supports LLM-based, encoder-based, and rule-based extraction regimes, balancing accuracy and inference cost through a common interface. Experiments show that Enoki remains competitive with strong claim-level systems while using fewer resources and achieves superior performance on fine-grained span- and entity-level localization. We also release EnokiQA, a dual-granularity dataset with aligned claim-level verification and span-level localization annotations.

cs.CL↗

arXiv2026

SoK: Evolution, Security, and Fundamental Properties of Transactional Systems

Transaction processing systems underpin modern commerce, finance, and critical infrastructure, yet their security has never been studied across the full evolutionary arc of these systems. Over five decades, transaction processing has progressed through four distinct generations, from centralized databases, to distributed databases, to blockchain and distributed ledger technologies (DLTs), finally to multi-context systems that span cyber-physical components under real-time constraints. Each generation has introduced new transaction types and new classes of vulnerabilities, yet security research remains fragmented by domain, and the foundational ACID transaction model has not been revisited to reflect the demands of contemporary systems. We classify 163 papers on transaction security by evolutionary generation, security focus, and relevant Common Weakness Enumeration (CWE) entries, and distill a curated set of 41 high-impact or seminal papers spanning all four generations. We make three principal contributions. First, we develop a four-generation evolutionary taxonomy that contextualizes each work within the broader trajectory of transaction processing. Second, we map each paper's security focus to CWE identifiers, providing a systems-oriented vocabulary for analyzing transaction-specific threats across otherwise siloed domains. Third, we demonstrate that the classical ACID properties are insufficient for modern transactional systems and introduce RANCID, extending ACID with Real-timeness (R) and N-many Contexts (N), as a property set for reasoning about the security and correctness of systems that must coordinate across heterogeneous contexts under timing constraints. Our systematization exposes a pronounced bias toward DLT security research at the expense of broader transactional security and identifies concrete open problems for the next generation of transaction processing systems.

cs.CR↗

arXiv2026

SoK: Self-Sovereign Digital Identities

Self-Sovereign Digital Identity (SSDI) enables individuals to control their own identity assertions and data, rather than relying on centralized or federated systems prone to large-scale data breaches. By eliminating centralized databases maintained by service providers and identity brokers, SSDIs offer enhanced security and privacy. However, adoption remains slow, and research in this area lacks systematization and uniformity. To address these gaps, we present a comprehensive systematization of knowledge on self-sovereign digital identities, with a primary focus on identifying the challenges that impede real-world adoption. We survey 80 academic and non-academic sources and identify six major challenges: (i) binding a single identity to one individual or organization, (ii) the absence of mature cryptographic and communication protocols, (iii) significant usability barriers, (iv) regulatory and oversight gaps, (v) bootstrapping to critical-mass adoption, and (vi) dependence on a permissionless, decentralized, yet singular infrastructure that may expose unforeseen vulnerabilities over time. We then analyze 47 scientific publications and find that the vast majority focus on blockchain-based solutions rather than generalized SSDI architectures. Additionally, we catalog 12 real-world, production-grade SSDI applications. Our evaluation of these solutions reveals that self-sovereignty is, in practice, a spectrum rather than a binary property. Finally, we explore the frontiers of SSDI by identifying major trends, open problems, and opportunities for future research. We hope this systematization will help advance the shift from centralized to self-sovereign digital identities in a disciplined and impactful way.

cs.CR↗

arXiv2026

Convergent-Beam X-ray Crystallography

Molecular and polymeric crystals show a wide range of functional properties that arise from the interplay between the atomic-scale structure of their constituent molecules and the organization of these molecules within the crystal lattice at macroscopic length scales. X-ray diffraction can provide structural information at these disparate length scales, but usually only through experiments that address one or the other of molecular (or unit-cell) structure versus crystal structure. Consequently, the accuracy of determined molecular or polymer structures may be limited by unaccounted crystal inhomogeneities of the crystal lattice and the characterization of crystalline materials might not reveal the underlying causes of crystal morphology. Here we introduce X-ray convergent-beam diffraction to obtain spatially-resolved structural information from crystals by projection topographic imaging. Using highly focusing X-ray multilayer Laue lenses, we show that Bragg reflections can be mapped into tomographic images of the crystal, for the characterization of strain and defects at high resolution. We demonstrate how the crystal morphology obtained this way can be accounted for when determining structure factors as a function of position in the crystal. The approach may assist in studies such as diffusion and binding in MOFS, protein-drug binding, crystal growth, and the mechanical responses of photo-reactive or thermally driven dynamic crystals.

cond-mat.mtrl-sci↗

arXiv2025

Eye of Judgement: Dissecting the Evaluation of Russian-speaking LLMs with POLLUX

We introduce POLLUX, a comprehensive open-source benchmark designed to evaluate the generative capabilities of large language models (LLMs) in Russian. Our main contribution is a novel evaluation methodology that enhances the interpretability of LLM assessment. For each task type, we define a set of detailed criteria and develop a scoring protocol where models evaluate responses and provide justifications for their ratings. This enables transparent, criteria-driven evaluation beyond traditional resource-consuming, side-by-side human comparisons. POLLUX includes a detailed, fine-grained taxonomy of 35 task types covering diverse generative domains such as code generation, creative writing, and practical assistant use cases, totaling 2,100 manually crafted and professionally authored prompts. Each task is categorized by difficulty (easy/medium/hard), with experts constructing the dataset entirely from scratch. We also release a family of LLM-as-a-Judge (7B and 32B) evaluators trained for nuanced assessment of generative outputs. This approach provides scalable, interpretable evaluation and annotation tools for model development, effectively replacing costly and less precise human judgments.

cs.CL↗

arXiv2025

Will It Still Be True Tomorrow? Multilingual Evergreen Question Classification to Improve Trustworthy QA

Large Language Models (LLMs) often hallucinate in question answering (QA) tasks. A key yet underexplored factor contributing to this is the temporality of questions -- whether they are evergreen (answers remain stable over time) or mutable (answers change). In this work, we introduce EverGreenQA, the first multilingual QA dataset with evergreen labels, supporting both evaluation and training. Using EverGreenQA, we benchmark 12 modern LLMs to assess whether they encode question temporality explicitly (via verbalized judgments) or implicitly (via uncertainty signals). We also train EG-E5, a lightweight multilingual classifier that achieves SoTA performance on this task. Finally, we demonstrate the practical utility of evergreen classification across three applications: improving self-knowledge estimation, filtering QA datasets, and explaining GPT-4o retrieval behavior.

cs.CL↗

arXiv2025

LLM-Independent Adaptive RAG: Let the Question Speak for Itself

Large Language Models~(LLMs) are prone to hallucinations, and Retrieval-Augmented Generation (RAG) helps mitigate this, but at a high computational cost while risking misinformation. Adaptive retrieval aims to retrieve only when necessary, but existing approaches rely on LLM-based uncertainty estimation, which remain inefficient and impractical. In this study, we introduce lightweight LLM-independent adaptive retrieval methods based on external information. We investigated 27 features, organized into 7 groups, and their hybrid combinations. We evaluated these methods on 6 QA datasets, assessing the QA performance and efficiency. The results show that our approach matches the performance of complex LLM-based methods while achieving significant efficiency gains, demonstrating the potential of external information for adaptive retrieval.

cs.CL↗

arXiv2025

Adaptive Retrieval Without Self-Knowledge? Bringing Uncertainty Back Home

Retrieval Augmented Generation (RAG) improves correctness of Question Answering (QA) and addresses hallucinations in Large Language Models (LLMs), yet greatly increase computational costs. Besides, RAG is not always needed as may introduce irrelevant information. Recent adaptive retrieval methods integrate LLMs' intrinsic knowledge with external information appealing to LLM self-knowledge, but they often neglect efficiency evaluations and comparisons with uncertainty estimation techniques. We bridge this gap by conducting a comprehensive analysis of 35 adaptive retrieval methods, including 8 recent approaches and 27 uncertainty estimation techniques, across 6 datasets using 10 metrics for QA performance, self-knowledge, and efficiency. Our findings show that uncertainty estimation techniques often outperform complex pipelines in terms of efficiency and self-knowledge, while maintaining comparable QA performance.

cs.CL↗

arXiv2024

Convergent-beam attosecond X-ray crystallography

Sub-angstrom spatial resolution of electron density coupled with sub-femtosecond temporal resolution is required to directly observe the dynamics of the electronic structure of a molecule after photoinitiation or some other ultrafast perturbation. Meeting this challenge, pushing the field of quantum crystallography to attosecond timescales, would bring insights into how the electronic and nuclear degrees of freedom couple, enable the study of quantum coherences involved in molecular dynamics, and ultimately enable these dynamics to be controlled. Here we propose to reach this realm by employing convergent-beam X-ray crystallography with high-power attosecond pulses from a hard-X-ray free-electron laser. We show that with dispersive optics, such as multilayer Laue lenses of high numerical aperture, it becomes possible to encode time into the resulting diffraction pattern with deep sub-femtosecond precision. Each snapshot diffraction pattern consists of Bragg streaks that can be mapped back to arrival times and positions of X-rays on the face of a crystal. This can span tens of femtoseconds, and can be finely sampled as we demonstrate experimentally. The approach brings several other advantages, such as an increase of the number of observable reflections in a snapshot diffraction pattern, all fully integrated, to improve the speed and accuracy of serial crystallography -- especially for crystals of small molecules.

physics.optics↗

arXiv2024

GigaPevt: Multimodal Medical Assistant

Building an intelligent and efficient medical assistant is still a challenging AI problem. The major limitation comes from the data modality scarceness, which reduces comprehensive patient perception. This demo paper presents the GigaPevt, the first multimodal medical assistant that combines the dialog capabilities of large language models with specialized medical models. Such an approach shows immediate advantages in dialog quality and metric performance, with a 1.18% accuracy improvement in the question-answering task.

cs.AI↗

arXiv2023

A Practical Survey on Emerging Threats from AI-driven Voice Attacks: How Vulnerable are Commercial Voice Control Systems?

The emergence of Artificial Intelligence (AI)-driven audio attacks has revealed new security vulnerabilities in voice control systems. While researchers have introduced a multitude of attack strategies targeting voice control systems (VCS), the continual advancements of VCS have diminished the impact of many such attacks. Recognizing this dynamic landscape, our study endeavors to comprehensively assess the resilience of commercial voice control systems against a spectrum of malicious audio attacks. Through extensive experimentation, we evaluate six prominent attack techniques across a collection of voice control interfaces and devices. Contrary to prevailing narratives, our results suggest that commercial voice control systems exhibit enhanced resistance to existing threats. Particularly, our research highlights the ineffectiveness of white-box attacks in black-box scenarios. Furthermore, the adversaries encounter substantial obstacles in obtaining precise gradient estimations during query-based interactions with commercial systems, such as Apple Siri and Samsung Bixby. Meanwhile, we find that current defense strategies are not completely immune to advanced attacks. Our findings contribute valuable insights for enhancing defense mechanisms in VCS. Through this survey, we aim to raise awareness within the academic community about the security concerns of VCS and advocate for continued research in this crucial area.

cs.CR↗

arXiv2023

DynamicFL: Balancing Communication Dynamics and Client Manipulation for Federated Learning

Federated Learning (FL) is a distributed machine learning (ML) paradigm, aiming to train a global model by exploiting the decentralized data across millions of edge devices. Compared with centralized learning, FL preserves the clients' privacy by refraining from explicitly downloading their data. However, given the geo-distributed edge devices (e.g., mobile, car, train, or subway) with highly dynamic networks in the wild, aggregating all the model updates from those participating devices will result in inevitable long-tail delays in FL. This will significantly degrade the efficiency of the training process. To resolve the high system heterogeneity in time-sensitive FL scenarios, we propose a novel FL framework, DynamicFL, by considering the communication dynamics and data quality across massive edge devices with a specially designed client manipulation strategy. \ours actively selects clients for model updating based on the network prediction from its dynamic network conditions and the quality of its training data. Additionally, our long-term greedy strategy in client selection tackles the problem of system performance degradation caused by short-term scheduling in a dynamic network. Lastly, to balance the trade-off between client performance evaluation and client manipulation granularity, we dynamically adjust the length of the observation window in the training process to optimize the long-term system efficiency. Compared with the state-of-the-art client selection scheme in FL, \ours can achieve a better model accuracy while consuming only 18.9\% -- 84.0\% of the wall-clock time. Our component-wise and sensitivity studies further demonstrate the robustness of \ours under various real-life scenarios.

cs.DC↗

arXiv2023

Security Defense For Smart Contracts: A Comprehensive Survey

The blockchain technology has been used for recording state transitions of smart contracts - decentralized applications that can be invoked through external transactions. Smart contracts gained popularity and accrued hundreds of billions of dollars in market capitalization in recent years. Unfortunately, like all other programs, smart contracts are prone to security vulnerabilities that have incurred multimillion-dollar damages over the past decade. As a result, many automated threat mitigation solutions have been proposed to counter the security issues of smart contracts. These threat mitigation solutions include various tools and methods that are challenging to compare. This survey develops a comprehensive classification taxonomy of smart contract threat mitigation solutions within five orthogonal dimensions: defense modality, core method, targeted contracts, input-output data mapping, and threat model. We classify 133 existing threat mitigation solutions using our taxonomy and confirm that the proposed five dimensions allow us to concisely and accurately describe any smart contract threat mitigation solution. In addition to learning what the threat mitigation solutions do, we also show how these solutions work by synthesizing their actual designs into a set of uniform workflows corresponding to the eight existing defense core methods. We further create an integrated coverage map for the known smart contract vulnerabilities by the existing threat mitigation solutions. Finally, we perform the evidence-based evolutionary analysis, in which we identify trends and future perspectives of threat mitigation in smart contracts and pinpoint major weaknesses of the existing methodologies. For the convenience of smart contract security developers, auditors, users, and researchers, we deploy a regularly updated comprehensive open-source online registry of threat mitigation solutions.

cs.CR↗

arXiv2023

TxT: Real-time Transaction Encapsulation for Ethereum Smart Contracts

Ethereum is a permissionless blockchain ecosystem that supports execution of smart contracts, the key enablers of decentralized finance (DeFi) and non-fungible tokens (NFT). However, the expressiveness of Ethereum smart contracts is a double-edged sword: while it enables blockchain programmability, it also introduces security vulnerabilities, i.e., the exploitable discrepancies between expected and actual behaviors of the contract code. To address these discrepancies and increase the vulnerability coverage, we propose a new smart contract security testing approach called transaction encapsulation. The core idea lies in the local execution of transactions on a fully-synchronized yet isolated Ethereum node, which creates a preview of outcomes of transaction sequences on the current state of blockchain. This approach poses a critical technical challenge -- the well-known time-of-check/time-of-use (TOCTOU) problem, i.e., the assurance that the final transactions will exhibit the same execution paths as the encapsulated test transactions. In this work, we determine the exact conditions for guaranteed execution path replicability of the tested transactions, and implement a transaction testing tool, TxT, which reveals the actual outcomes of Ethereum transactions. To ensure the correctness of testing, TxT deterministically verifies whether a given sequence of transactions ensues an identical execution path on the current state of blockchain. We analyze over 1.3 billion Ethereum transactions and determine that 96.5% of them can be verified by TxT. We further show that TxT successfully reveals the suspicious behaviors associated with 31 out of 37 vulnerabilities (83.8% coverage) in the smart contract weakness classification (SWC) registry. In comparison, the vulnerability coverage of all the existing defense approaches combined only reaches 40.5%.

cs.CR↗

arXiv2022

Is Rust C++-fast? Benchmarking System Languages on Everyday Routines

Rust is a relatively new system programming language that has been experiencing a rapid adoption in the past 10 years. Rust incorporates a memory ownership model enforced at a compile time. Since this model involves zero runtime overhead, programs written in Rust are not only memory-safe but also fast, leading to performance comparable to C and C++. Multiple existing benchmarks comparing the performance of Rust with other languages focus on rarely used superficial algorithms, leading to somewhat inconclusive results. In this work, we conduct a comparative performance benchmark of Rust and C++ using commonly used algorithms and data structures rather than exotic ones. Our evaluation shows that the overall performance of Rust is similar to C++, with only minor disadvantage. We also demonstrate that in some Rust routines are slightly faster than the ones of C++.

cs.PL↗

arXiv2022

Et tu, Blockchain? Outsmarting Smart Contracts via Social Engineering

We reveal six zero-day social engineering attacks in Ethereum, and subdivide them into two classes: Address Manipulation and Homograph. We demonstrate the attacks by embedding them in source codes of five popular smart contracts with combined market capitalization of over \$29 billion, and show that the attacks have the ability to remain dormant during the testing phase and activate only after production deployment. We analyze 85,656 open source smart contracts and find 1,027 contracts that can be directly used for performing social engineering attacks. For responsible disclosure, we contact seven smart contract security firms. In the spirit of open research, we make the source codes of the attack benchmark, tools, and datasets available to the public.

cs.CR↗

arXiv2022

Decentralization Paradox: A Study of Hegemonic and Risky ERC-20 Tokens

In this work, we explore the class of Ethereum smart contracts called the administrated ERC20 tokens. We demonstrate that these contracts are more owner-controlled and less safe than the services they try to disrupt, such as banks and centralized online payment systems. We develop a binary classifier for identification of administrated ERC20 tokens, and conduct extensive data analysis, which reveals that nearly 9 out of 10 ERC20 tokens on Ethereum are administrated, and thereby unsafe to engage with even under the assumption of trust towards their owners. We design and implement SafelyAdministrated - a Solidity abstract class that safeguards users of administrated ERC20 tokens from adversarial attacks or frivolous behavior of the tokens' owners.

cs.CR↗

arXiv2022

SuperVoice: Text-Independent Speaker Verification Using Ultrasound Energy in Human Speech

Voice-activated systems are integrated into a variety of desktop, mobile, and Internet-of-Things (IoT) devices. However, voice spoofing attacks, such as impersonation and replay attacks, in which malicious attackers synthesize the voice of a victim or simply replay it, have brought growing security concerns. Existing speaker verification techniques distinguish individual speakers via the spectrographic features extracted from an audible frequency range of voice commands. However, they often have high error rates and/or long delays. In this paper, we explore a new direction of human voice research by scrutinizing the unique characteristics of human speech at the ultrasound frequency band. Our research indicates that the high-frequency ultrasound components (e.g. speech fricatives) from 20 to 48 kHz can significantly enhance the security and accuracy of speaker verification. We propose a speaker verification system, SUPERVOICE that uses a two-stream DNN architecture with a feature fusion mechanism to generate distinctive speaker models. To test the system, we create a speech dataset with 12 hours of audio (8,950 voice samples) from 127 participants. In addition, we create a second spoofed voice dataset to evaluate its security. In order to balance between controlled recordings and real-world applications, the audio recordings are collected from two quiet rooms by 8 different recording devices, including 7 smartphones and an ultrasound microphone. Our evaluation shows that SUPERVOICE achieves 0.58% equal error rate in the speaker verification task, it only takes 120 ms for testing an incoming utterance, outperforming all existing speaker verification systems. Moreover, within 91 ms processing time, SUPERVOICE achieves 0% equal error rate in detecting replay attacks launched by 5 different loudspeakers.

cs.SD↗