arXiv subjects

Radu State

Publications and source records attributed to Radu State.

At least 37 records · Page 2Linked to original sources

arXiv2026

FedRandom: Sampling Consistent and Accurate Contribution Values in Federated Learning

Federated Learning is a privacy-preserving decentralized approach for Machine Learning tasks. In industry deployments characterized by a limited number of entities possessing abundant data, the significance of a participant's role in shaping the global model becomes pivotal given that participation in a federation incurs costs, and participants may expect compensation for their involvement. Additionally, the contributions of participants serve as a crucial means to identify and address potential malicious actors and free-riders. However, fairly assessing individual contributions remains a significant hurdle. Recent works have demonstrated a considerable inherent instability in contribution estimations across aggregation strategies. While employing a different strategy may offer convergence benefits, this instability can have potentially harming effects on the willingness of participants in engaging in the federation. In this work, we introduce FedRandom, a novel mitigation technique to the contribution instability problem. Tackling the instability as a statistical estimation problem, FedRandom allows us to generate more samples than when using regular FL strategies. We show that these additional samples provide a more consistent and reliable evaluation of participant contributions. We demonstrate our approach using different data distributions across CIFAR-10, MNIST, CIFAR-100 and FMNIST and show that FedRandom reduces the overall distance to the ground truth by more than a third in half of all evaluated scenarios, and improves stability in more than 90% of cases.

cs.LG↗

arXiv2025

Temporal-Spatial Tubelet Embedding for Cloud-Robust MSI Reconstruction using MSI-SAR Fusion: A Multi-Head Self-Attention Video Vision Transformer Approach

Cloud cover in multispectral imagery (MSI) significantly hinders early-season crop mapping by corrupting spectral information. Existing Vision Transformer(ViT)-based time-series reconstruction methods, like SMTS-ViT, often employ coarse temporal embeddings that aggregate entire sequences, causing substantial information loss and reducing reconstruction accuracy. To address these limitations, a Video Vision Transformer (ViViT)-based framework with temporal-spatial fusion embedding for MSI reconstruction in cloud-covered regions is proposed in this study. Non-overlapping tubelets are extracted via 3D convolution with constrained temporal span $(t=2)$, ensuring local temporal coherence while reducing cross-day information degradation. Both MSI-only and SAR-MSI fusion scenarios are considered during the experiments. Comprehensive experiments on 2020 Traill County data demonstrate notable performance improvements: MTS-ViViT achieves a 2.23\% reduction in MSE compared to the MTS-ViT baseline, while SMTS-ViViT achieves a 10.33\% improvement with SAR integration over the SMTS-ViT baseline. The proposed framework effectively enhances spectral reconstruction quality for robust agricultural monitoring.

cs.CV↗

arXiv2025

Uncovering Zero-Shot Generalization Gaps in Time-Series Foundation Models Using Real-World Videos

Recent research on time-series foundation models (TSFMs) has underscored the scarcity of real-world data, often supplemented with synthetic sources in existing datasets, whose generalizability remains however debated. As such, in this work, we propose a novel benchmarking approach: in particular, we aim at building a curated dataset reflecting real world physical temporal dynamics, extracting temporal signals from real-world videos using optical flow. As such, we introduce REAL-V-TSFM, a novel dataset designed to capture rich and diverse time series derived from real-world videos. Experimental results on state-of-the-art TSFMs under zero-shot forecasting show that, despite strong performance on conventional benchmarks, these models exhibit performance degradation on the proposed dataset, suggesting limited generalizability to novel datasets. These findings underscore the need for novel approaches to acquiring time series data and highlight the lack of universality in recent TSFMs, while further validating the effectiveness of our video-based time series data extraction pipeline.

cs.AI↗

arXiv2025

NegBLEURT Forest: Leveraging Inconsistencies for Detecting Jailbreak Attacks

Jailbreak attacks designed to bypass safety mechanisms pose a serious threat by prompting LLMs to generate harmful or inappropriate content, despite alignment with ethical guidelines. Crafting universal filtering rules remains difficult due to their inherent dependence on specific contexts. To address these challenges without relying on threshold calibration or model fine-tuning, this work introduces a semantic consistency analysis between successful and unsuccessful responses, demonstrating that a negation-aware scoring approach captures meaningful patterns. Building on this insight, a novel detection framework called NegBLEURT Forest is proposed to evaluate the degree of alignment between outputs elicited by adversarial prompts and expected safe behaviors. It identifies anomalous responses using the Isolation Forest algorithm, enabling reliable jailbreak detection. Experimental results show that the proposed method consistently achieves top-tier performance, ranking first or second in accuracy across diverse models using the crafted dataset, while competing approaches exhibit notable sensitivity to model and data variations.

cs.CR↗

arXiv2025

Blockly2Hooks: Smart Contracts for Everyone with the XRP Ledger and Google Blockly

Recent technologies such as inter-ledger payments, non-fungible tokens, and smart contracts are all fruited from the ongoing development of Distributed Ledger Technologies. The foreseen trend is that they will play an increasingly visible role in daily life, which will have to be backed by appropriate operational resources. For example, due to increasing demand, smart contracts could soon face a shortage of knowledgeable users and tools to handle them in practice. Widespread smart contract adoption is currently limited by security, usability and costs aspects. Because of a steep learning curve, the handling of smart contracts is currently performed by specialised developers mainly, and most of the research effort is focusing on smart contract security, while other aspects like usability being somewhat neglected. Specific tools would lower the entry barrier, enabling interested non-experts to create smart contracts. In this paper we designed, developed and tested Blockly2Hooks, a solution towards filling this gap even in challenging scenarios such as when the smart contracts are written in an advanced language like C. With the XRP Ledger as a concrete working case, Blockly2Hooks helps interested non-experts from the community to learn smart contracts easily and adopt the technology, through leveraging well-proven teaching methodologies like Visual Programming Languages, and more specifically, the Blockly Visual Programming library from Google. The platform was developed and tested and the results are promising to make learning smart contract development smoother.

cs.CR↗

arXiv2025

To Squelch or not to Squelch: Enabling Improved Message Dissemination on the XRP Ledger

With the large increase in the adoption of blockchain technologies, their underlying peer-to-peer networks must also scale with the demand. In this context, previous works highlighted the importance of ensuring efficient and resilient communication for the underlying consensus and replication mechanisms. However, they were mainly focused on mainstream, Proof-of-Work-based Distributed Ledger Technologies like Bitcoin or Ethereum. In this paper, the problem is investigated in the context of consensus-validation based blockchains, like the XRP Ledger. The latter relies on a Federated Byzantine Agreement (FBA) consensus mechanism which is proven to have a good scalability in regards to transaction throughput. However, it is known that significant increases in the size of the XRP Ledger network would be challenging to achieve. The main reason is the flooding mechanism used to disseminate the messages related to the consensus protocol, which creates many duplicates in the network. Squelching is a recent solution proposed for limiting this duplication, however, it was never evaluated quantitatively in real-life scenarios involving the XRPL production network. In this paper, our aim is to assess this mechanism using a real-life controllable testbed and the XRPL production network, to assess its benefit and compare it to alternative solutions relying on Named Data Networking and on a gossip-based approach.

cs.NI↗

arXiv2025

HalluGuard: Evidence-Grounded Small Reasoning Models to Mitigate Hallucinations in Retrieval-Augmented Generation

Large Language Models (LLMs) excel in many NLP tasks but remain prone to hallucinations, limiting trust in real-world applications. We present HalluGuard, a 4B-parameter Small Reasoning Model (SRM) for mitigating hallucinations in Retrieval-Augmented Generation (RAG). HalluGuard classifies document-claim pairs as grounded or hallucinated and produces evidence-grounded justifications for transparency. Our approach combines (i) a domain-agnostic synthetic dataset derived from FineWeb and refined through multi-stage curation and data reformation, (ii) synthetic grounded and hallucinated claims, and (iii) preference-based fine-tuning with Odds Ratio Preference Optimization to distill large-model reasoning into a smaller backbone. On the RAGTruth subset of the LLM-AggreFact benchmark, HalluGuard achieves 84.0% balanced accuracy (BAcc), rivaling specialized models, MiniCheck (7B; 84.0%) and Granite Guardian 3.3 (8B; 82.2%) while using roughly half their parameters. Over the full benchmark it reaches 75.7% BAcc, matching larger general-purpose LLMs such as GPT-4o (75.9%). We will release HalluGuard and datasets under Apache 2.0 upon acceptance.

cs.CL↗

arXiv2025

Is Small Language Model the Silver Bullet to Low-Resource Languages Machine Translation?

Low-resource languages (LRLs) lack sufficient linguistic resources and are underrepresented in benchmark datasets, resulting in persistently lower translation quality than high-resource languages, especially in privacy-sensitive and resource-limited contexts. Firstly, this study systematically evaluates state-of-the-art smaller Large Language Models in 200 languages using the FLORES-200 benchmark, highlighting persistent deficiencies and disparities in the translation of LRLs. To mitigate these limitations, we investigate knowledge distillation from large pre-trained teacher models to Small Language Models (SLMs) through supervised fine-tuning. The results show substantial improvements; for example, the translation performance of English to Luxembourgish (EN to LB), measured by the LLM-as-a-Judge score, increases from 0.36 to 0.89 in the validation set for Llama-3.2-3B. We further investigate various fine-tuning configurations and tasks to clarify the trade-offs between data scale and training efficiency, verify that the model retains its general capabilities without significant catastrophic forgetting after training, and explore the distillation benefits to other LRLs on SLMs (Khasi, Assamese, and Ukrainian). In general, this work exposes the limitations and fairness issues of current SLMs in LRL translation and systematically explores the potential of using the distillation of knowledge from large to small models, offering practical, empirically grounded recommendations to improve LRL translation systems

cs.CL↗

arXiv2025

WallStreetFeds: Client-Specific Tokens as Investment Vehicles in Federated Learning

Federated Learning (FL) is a collaborative machine learning paradigm which allows participants to collectively train a model while training data remains private. This paradigm is especially beneficial for sectors like finance, where data privacy, security and model performance are paramount. FL has been extensively studied in the years following its introduction, leading to, among others, better performing collaboration techniques, ways to defend against other clients trying to attack the model, and contribution assessment methods. An important element in for-profit Federated Learning is the development of incentive methods to determine the allocation and distribution of rewards for participants. While numerous methods for allocation have been proposed and thoroughly explored, distribution frameworks remain relatively understudied. In this paper, we propose a novel framework which introduces client-specific tokens as investment vehicles within the FL ecosystem. Our framework aims to address the limitations of existing incentive schemes by leveraging a decentralized finance (DeFi) platform and automated market makers (AMMs) to create a more flexible and scalable reward distribution system for participants, and a mechanism for third parties to invest in the federation learning process.

cs.LG↗

arXiv2025

Small Language Models in the Real World: Insights from Industrial Text Classification

With the emergence of ChatGPT, Transformer models have significantly advanced text classification and related tasks. Decoder-only models such as Llama exhibit strong performance and flexibility, yet they suffer from inefficiency on inference due to token-by-token generation, and their effectiveness in text classification tasks heavily depends on prompt quality. Moreover, their substantial GPU resource requirements often limit widespread adoption. Thus, the question of whether smaller language models are capable of effectively handling text classification tasks emerges as a topic of significant interest. However, the selection of appropriate models and methodologies remains largely underexplored. In this paper, we conduct a comprehensive evaluation of prompt engineering and supervised fine-tuning methods for transformer-based text classification. Specifically, we focus on practical industrial scenarios, including email classification, legal document categorization, and the classification of extremely long academic texts. We examine the strengths and limitations of smaller models, with particular attention to both their performance and their efficiency in Video Random-Access Memory (VRAM) utilization, thereby providing valuable insights for the local deployment and application of compact models in industrial settings.

cs.CL↗

arXiv2025

Exploring the Impact of Temperature on Large Language Models:Hot or Cold?

The sampling temperature, a critical hyperparameter in large language models (LLMs), modifies the logits before the softmax layer, thereby reshaping the distribution of output tokens. Recent studies have challenged the Stochastic Parrots analogy by demonstrating that LLMs are capable of understanding semantics rather than merely memorizing data and that randomness, modulated by sampling temperature, plays a crucial role in model inference. In this study, we systematically evaluated the impact of temperature in the range of 0 to 2 on data sets designed to assess six different capabilities, conducting statistical analyses on open source models of three different sizes: small (1B--4B), medium (6B--13B), and large (40B--80B). Our findings reveal distinct skill-specific effects of temperature on model performance, highlighting the complexity of optimal temperature selection in practical applications. To address this challenge, we propose a BERT-based temperature selector that takes advantage of these observed effects to identify the optimal temperature for a given prompt. We demonstrate that this approach can significantly improve the performance of small and medium models in the SuperGLUE datasets. Furthermore, our study extends to FP16 precision inference, revealing that temperature effects are consistent with those observed in 4-bit quantized models. By evaluating temperature effects up to 4.0 in three quantized models, we find that the Mutation Temperature -- the point at which significant performance changes occur -- increases with model size.

cs.CL↗

arXiv2025

On the Volatility of Shapley-Based Contribution Metrics in Federated Learning

Federated learning (FL) is a collaborative and privacy-preserving Machine Learning paradigm, allowing the development of robust models without the need to centralize sensitive data. A critical challenge in FL lies in fairly and accurately allocating contributions from diverse participants. Inaccurate allocation can undermine trust, lead to unfair compensation, and thus participants may lack the incentive to join or actively contribute to the federation. Various remuneration strategies have been proposed to date, including auction-based approaches and Shapley-value-based methods, the latter offering a means to quantify the contribution of each participant. However, little to no work has studied the stability of these contribution evaluation methods. In this paper, we evaluate participant contributions in federated learning using gradient-based model reconstruction techniques with Shapley values and compare the round-based contributions to a classic data contribution measurement scheme. We provide an extensive analysis of the discrepancies of Shapley values across a set of aggregation strategies and examine them on an overall and a per-client level. We show that, between different aggregation techniques, Shapley values lead to unstable reward allocations among participants. Our analysis spans various data heterogeneity distributions, including independent and identically distributed (IID) and non-IID scenarios.

cs.LG↗

arXiv2024

LongKey: Keyphrase Extraction for Long Documents

In an era of information overload, manually annotating the vast and growing corpus of documents and scholarly papers is increasingly impractical. Automated keyphrase extraction addresses this challenge by identifying representative terms within texts. However, most existing methods focus on short documents (up to 512 tokens), leaving a gap in processing long-context documents. In this paper, we introduce LongKey, a novel framework for extracting keyphrases from lengthy documents, which uses an encoder-based language model to capture extended text intricacies. LongKey uses a max-pooling embedder to enhance keyphrase candidate representation. Validated on the comprehensive LDKP datasets and six diverse, unseen datasets, LongKey consistently outperforms existing unsupervised and language model-based keyphrase extraction methods. Our findings demonstrate LongKey's versatility and superior performance, marking an advancement in keyphrase extraction for varied text lengths and domains.

cs.CL↗

arXiv2024

Cross Domain Early Crop Mapping using CropSTGAN

Driven by abundant satellite imagery, machine learning-based approaches have recently been promoted to generate high-resolution crop cultivation maps to support many agricultural applications. One of the major challenges faced by these approaches is the limited availability of ground truth labels. In the absence of ground truth, existing work usually adopts the "direct transfer strategy" that trains a classifier using historical labels collected from other regions and then applies the trained model to the target region. Unfortunately, the spectral features of crops exhibit inter-region and inter-annual variability due to changes in soil composition, climate conditions, and crop progress, the resultant models perform poorly on new and unseen regions or years. Despite recent efforts, such as the application of the deep adaptation neural network (DANN) model structure in the deep adaptation crop classification network (DACCN), to tackle the above cross-domain challenges, their effectiveness diminishes significantly when there is a large dissimilarity between the source and target regions. This paper introduces the Crop Mapping Spectral-temporal Generative Adversarial Neural Network (CropSTGAN), a novel solution for cross-domain challenges, that doesn't require target domain labels. CropSTGAN learns to transform the target domain's spectral features to those of the source domain, effectively bridging large dissimilarities. Additionally, it employs an identity loss to maintain the intrinsic local structure of the data. Comprehensive experiments across various regions and years demonstrate the benefits and effectiveness of the proposed approach. In experiments, CropSTGAN is benchmarked against various state-of-the-art (SOTA) methods. Notably, CropSTGAN significantly outperforms these methods in scenarios with large data distribution dissimilarities between the target and source domains.

cs.CV↗

arXiv2023

XRP-NDN Overlay: Improving the Communication Efficiency of Consensus-Validation based Blockchains with an NDN Overlay

With the growing adoption of Distributed Ledger Technologies and the subsequent scaling of these networks, there is an inherent need for efficient and resilient communication used by the underlying consensus and replication mechanisms. While resilient and efficient communication is one of the main pillars of an efficient blockchain network as a whole, the Distributed Ledger Technology is still relatively new and the task of scaling these networks has come with its own challenges towards ensuring these goals. New content distribution concepts like Information Centric Networking, of which Named Data Networking is a worthy example, create new possibilities towards achieving this goal, through in-network caching or built-in native multicasting, for example. We present and evaluate XRP-NDN Overlay, a solution for increasing the communication efficiency for consensus-validation based blockchains like the XRP Ledger. We experiment by sending the XRP Ledger consensus messages over different Named Data Networking communication models and prove that our chosen model lowers the number of messages at node level to minimum necessary, while maintaining or improving blockchain performance by leveraging the possibilities offered by an overlay such as specific communication mechanisms.

cs.NI↗

arXiv2023

Topology Analysis of the XRP Ledger

XRP Ledger is one of the oldest, well-established blockchains. Despite the popularity of the XRP Ledger, little is known about its underlying peer-to-peer network. The structural properties of a network impact its efficiency, security and robustness. We aim to close the knowledge gap by providing a detailed analysis of the XRP overlay network. In this paper we examine the graph-theoretic properties of the XRP Ledger peer-to-peer network and its temporal characteristics. We crawl the XRP Ledger over two months and collect 1,290 unique network snapshots. We uncover a small group of nodes that act as a networking backbone. In addition, we observe a high network churn, with a third of the nodes changing every five days. Our findings have strong implications for the resilience and safety of the XRP Ledger.

cs.CE↗

arXiv2022

A Flash(bot) in the Pan: Measuring Maximal Extractable Value in Private Pools

The rise of Ethereum has lead to a flourishing decentralized marketplace that has, unfortunately, fallen victim to frontrunning and Maximal Extractable Value (MEV) activities, where savvy participants game transaction orderings within a block for profit. One popular solution to address such behavior is Flashbots, a private pool with infrastructure and design goals aimed at eliminating the negative externalities associated with MEV. While Flashbots has established laudable goals to address MEV behavior, no evidence has been provided to show that these goals are achieved in practice. In this paper, we measure the popularity of Flashbots and evaluate if it is meeting its chartered goals. We find that (1) Flashbots miners account for over 99.9% of the hashing power in the Ethereum network, (2) powerful miners are making more than $2\times$ what they were making prior to using Flashbots, while non-miners' slice of the pie has shrunk commensurately, (3) mining is just as centralized as it was prior to Flashbots with more than 90% of Flashbots blocks coming from just two miners, and (4) while more than 80% of MEV extraction in Ethereum is happening through Flashbots, 13.2% is coming from other private pools.

cs.CR↗

arXiv2022

Elysium: Context-Aware Bytecode-Level Patching to Automatically Heal Vulnerable Smart Contracts

Fixing bugs is easiest by patching source code. However, source code is not always available: only 0.3% of the ~49M smart contracts that are currently deployed on Ethereum have their source code publicly available. Moreover, since contracts may call functions from other contracts, security flaws in closed-source contracts may affect open-source contracts as well. However, current state-of-the-art approaches that operate on closed-source contracts (i.e., EVM bytecode), such as EVMPatch and SmartShield, make use of purely hard-coded templates that leverage fix patching patterns. As a result, they cannot dynamically adapt to the bytecode that is being patched, which severely limits their flexibility and scalability. For instance, when patching integer overflows using hard-coded templates, a particular patch template needs to be employed as the bounds to be checked are different for each integer size. In this paper, we propose Elysium, a scalable approach towards automatic smart contract repair at the bytecode level. Elysium combines template-based and semantic-based patching by inferring context information from bytecode. Elysium is currently able to patch 7 different types of vulnerabilities in smart contracts automatically and can easily be extended with new templates and new bug-finding tools. We evaluate its effectiveness and correctness using 3 different datasets by replaying more than 500K transactions on patched contracts. We find that Elysium outperforms existing tools by patching at least 30% more contracts correctly. Finally, we also compare the overhead of Elysium in terms of deployment and transaction cost. In comparison to other tools, we find that generally Elysium minimizes the runtime cost (i.e., transaction cost) up to a factor of 1.7, for only a marginally higher deployment cost, where deployment cost is a one-time cost as compared to the runtime cost.

cs.CR↗