Multiple Tokens Research Articles

VCC-DiffNet: Visual Conditional Control Diffusion Network for Remote Sensing Image Captioning

Pioneering remote sensing image captioning (RSIC) works use autoregressive decoding for fluent and coherent sentences but suffer from high latency and high computation costs. In contrast, non-autoregressive approaches improve inference speed by predicting multiple tokens simultaneously, though at the cost of performance due to a lack of sequential dependencies. Recently, diffusion model-based non-autoregressive decoding has shown promise in natural image captioning with iterative refinement, but its effectiveness is limited by the intrinsic characteristics of remote sensing images, which complicate robust input construction and affect the description accuracy. To overcome these challenges, we propose an innovative diffusion model for RSIC, named the Visual Conditional Control Diffusion Network (VCC-DiffNet). Specifically, we propose a Refined Multi-scale Feature Extraction (RMFE) module to extract the discernible visual context features of RSIs as input of the diffusion model-based non-autoregressive decoder to conditionally control a multi-step denoising process. Furthermore, we propose an Interactive Enhanced Decoder (IE-Decoder) utilizing dual image–description interactions to generate descriptions finely aligned with the image content. Experiments conducted on four representative RSIC datasets demonstrate that our non-autoregressive VCC-DiffNet performs comparably to, or even better than, popular autoregressive baselines in classic metrics, achieving around an 8.22× speedup in Sydney-Captions, an 11.61× speedup in UCM-Captions, a 15.20× speedup in RSICD, and an 8.13× speedup in NWPU-Captions.

Improving biomedical entity linking for complex entity mentions with LLM-based text simplification.

Large amounts of important medical information are captured in free-text documents in biomedical research and within healthcare systems, which can be made accessible through natural language processing (NLP). A key component in most biomedical NLP pipelines is entity linking, i.e. grounding textual mentions of named entities to a reference of medical concepts, usually derived from a terminology system, such as the Systematized Nomenclature of Medicine Clinical Terms. However, complex entity mentions, spanning multiple tokens, are notoriously hard to normalize due to the difficulty of finding appropriate candidate concepts. In this work, we propose an approach to preprocess such mentions for candidate generation, building upon recent advances in text simplification with generative large language models. We evaluate the feasibility of our method in the context of the entity linking track of the BioCreative VIII SympTEMIST shared task. We find that instructing the latest Generative Pre-trained Transformer model with a few-shot prompt for text simplification results in mention spans that are easier to normalize. Thus, we can improve recall during candidate generation by 2.9percentage points compared to our baseline system, which achieved the best score in the original shared task evaluation. Furthermore, we show that this improvement in recall can be fully translated into top-1 accuracy through careful initialization of a subsequent reranking model. Our best system achieves an accuracy of 63.6% on the SympTEMIST test set. The proposed approach has been integrated into the open-source xMEN toolkit, which is available online via https://github.com/hpi-dhc/xmen.

MT2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}AD: multi-layer temporal transaction anomaly detection in ethereum networks with GNN

In recent years, a surge of criminal activities with cross-cryptocurrency trades have emerged in Ethereum, the second-largest public blockchain platform. Most of the existing anomaly detection methods utilize the traditional machine learning with feature engineering or graph representation learning technique to capture the information in transaction network. However, these methods either ignore the timestamp information and the transaction flow direction information in transaction network or only consider single transaction network, the cross-cryptocurrency trading patterns in Ethereum are usually ignored. In this paper, we introduce a Multi-layer Temporal Transaction Anomaly Detection (MT2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}AD) model in Ethereum network with graph neural network. Specifically, for a given Ethereum token transaction network, we first extract its initial features including the structure subgraph and edge’s feature. Then, we model the temporal information in subgraph as a series of network snapshots according to the timestamp on each edge and time window. To capture the cross-cryptocurrency trading patterns, we combine the snapshots from multiple token transactions at a given timestamp, and we consider it as a new combined graph. We further use the graph convolution encoder with attention mechanism and pooling operation on this new graph to obtain the graph-level embedding, and we transform the anomaly detection on dynamic multi-layer Ethereum transaction networks as a graph classification task with these graph-level embeddings. MT2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}AD can integrate the transaction structure feature, edge’s feature and cross-cryptocurrency trading patterns into a framework to perform the anomaly detection with graph neural networks. Experiments on three real-world multi-layer transaction networks show that the proposed MT2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^2$$\\end{document}AD (0.8789 Precision, 0.9375 Recall, 0.4987 FbMacro and 0.9351 FbWeighted) can achieve the best performance on most evaluation metrics in comparison with some competing approaches, and the effectiveness in consideration of multiple tokens is also demonstrated.

Prompt Learning with Structured Semantic Knowledge Makes Pre-Trained Language Models Better

Pre-trained language models with structured semantic knowledge have demonstrated remarkable performance in a variety of downstream natural language processing tasks. The typical methods of integrating knowledge are designing different pre-training tasks and training from scratch, which requires high-end hardware, massive storage resources, and long computing times. Prompt learning is an effective approach to tuning language models for specific tasks, and it can also be used to infuse knowledge. However, most prompt learning methods accept one token as the answer, instead of multiple tokens. To tackle this problem, we propose the long-answer prompt learning method (KLAPrompt), with three different long-answer strategies, to incorporate semantic knowledge into pre-trained language models, and we compare the performance of these three strategies through experiments. We also explore the effectiveness of the KLAPrompt method in the medical field. Additionally, we generate a word sense prediction dataset (WSP) based on the Xinhua Dictionary and a disease and category prediction dataset (DCP) based on MedicalKG. Experimental results show that discrete answers with the answer space partitioning strategy achieve the best results, and introducing structured semantic information can consistently improve language modeling and downstream tasks.

What Makes for Good Tokenizers in Vision Transformer?

The architecture of transformers, which recently witness booming applications in vision tasks, has pivoted against the widespread convolutional paradigm. Relying on the tokenization process that splits inputs into multiple tokens, transformers are capable of extracting their pairwise relationships using self-attention. While being the stemming building block of transformers, what makes for a good tokenizer has not been well understood in computer vision. In this work, we investigate this uncharted problem from an information trade-off perspective. In addition to unifying and understanding existing structural modifications, our derivation leads to better design strategies for vision tokenizers. The proposed Modulation across Tokens (MoTo) incorporates inter-token modeling capability through normalization. Furthermore, a regularization objective TokenProp is embraced in the standard training regime. Through extensive experiments on various transformer architectures, we observe both improved performance and intriguing properties of these two plug-and-play designs with negligible computational overhead. These observations further indicate the importance of the commonly-omitted designs of tokenizers in vision transformer.

Tracking vision transformer with class and regression tokens

Aim of this work is to introduce a novel visual object tracking model based on siamese network and vision transformer. Tracking is performed by multiple tokens exploiting the learning and memorization capabilities of the vision transformers. Therefore, the tracking problem is divided into multiple sub-tasks and experiments by using multiple tokens for learning each individual sub-task. This makes possible to learn a robust characterization of the problem with an explainable architecture, understanding the motivation of the choice that the neural network does. This is due to the attention in the transformer that uses the representational capacity of tokens that allows one to identify, simply with respect to different architectures and methodologies, where all the interest is focused. Several experiments are performed on benchmark data proving to be among the most performing trackers compared with the state of the art in explainability, precision, robustness and speed.

A marker collaborating model for entity and relation extraction

Recognizing entities and extracting their relations are significant tasks for information extraction. Many recent works focus on leveraging entity markers for better span (pair) representations, which have achieved promising performance. Nevertheless, existing works have two shortcomings: (1) previous markers are randomly embedded into distributed representations, which ignores semantic information relevant to the targeted tokens that we want to highlight. (2) most works simply implant markers into a sentence, which is weak to encode the interrelation between multiple tokens. In this work, we propose a marker collaborating model for entity and relation extraction, which is composed of two modules. The first module is semantic marker module, where two kinds of semantic markers are proposed to leverage semantic features from the PLMs. The second module is collaborative marker attention module, where a marker-oriented directional attention is proposed to model the interaction between multiple tokens. Our model yields SOTA performance on three standard benchmarks (ACE04, ACE05 and SciERC).

Token Multiplicity in Reversing Petri Nets Under the Individual Token Interpretation

Reversing Petri nets (RPNs) have recently been proposed as a net-basedapproach to model causal and out-of-causal order reversibility. They are based on the notion of individual tokens that can be connected together via bonds. In this paper we extend RPNs by allowing multiple tokens of the same type to exist within a net based on the individual token interpretation of Petri nets. According to this interpretation, tokens of the same type are distinguished via their causal path. We develop a causal semantics of the model and we prove that the expressive power of RPNs with multiple tokens is equivalent to that of RPNs with single tokens by establishing an isomporphism between the Labelled Transition Systems (LTSs) capturing the reachable parts of the respective RPN models.

Abstract processes in the absence of conflicts in general place/transition systems

Goltz and Reisig generalised Petri's concept of processes of one-safe Petri nets to general nets where places carry multiple tokens. BD-processes are equivalence classes of Goltz-Reisig processes connected through the swapping transformation of Best and Devillers; they can be considered as an alternative representation of runs of nets. Here we present an order respecting bijection between the BD-processes and the FS-processes of a countable net, the latter being defined—in an analogous way—as equivalence classes of firing sequences. Using this, we show that a countable net without binary conflicts has a (unique) largest BD-process.

A collective interpretation semantics for reversing Petri nets

Reversing Petri nets (RPNs) have recently been proposed as a net-based approach to model causal and out-of-causal order reversibility. They are based on the notion of individual tokens that can be connected together via bonds. In this paper we extend RPNs by allowing multiple tokens of the same type to exist within a net and we implement a notion of local reversibility for the model based on the collective token interpretation of Petri nets. The resulting framework allows us to model a form of out-of-causal-order reversibility, a type of reversibility that arises most notably in biological systems. Furthermore, it is lightweight in the sense that it requires no global control and no history needs to be preserved. Furthermore, we provide a mechanism for controlling reversibility by associating transitions with conditions, which guard their execution. We demonstrate the usefulness of the framework by modelling the functioning of the sodium-potassium pump, a system where transition execution is governed by activation conditions and featuring reversible behaviour.

Conformance evaluation of the top‐100 Ethereum token smart contracts with Ethereum Request for Comment‐20 functional specifications

This study is an attempt to evaluate the compliance problem regarding the current status of the establishment of agreed-upon interfaces for tokens, which are the most important type of applications of the Ethereum blockchain. The authors analysed the existing application programming interface standard interface called Ethereum Request for Comment 20 (ERC-20), proposed functional ERC-20 specifications, and evaluated the real-world token smart contracts that have been deployed and used for transactions on a blockchain. For the evaluation, the authors developed an automatic tool with a test suite based on the proposed functional specifications. The authors' experiment revealed that no top-100 market capitalization token smart contracts turned out to agree perfectly based on the functional specifications. In addition, the authors found the de facto standard behaviour deduced through testing with the functional specifications and suggested a possible countermeasure for the compliance problem in dealing with multiple tokens together.

An Eye-Tracking Study on EFL Learners’ Subject-Verb Agreement Processing

Eye-tracking technology enables us to collect and analyze a wide range of eye movement measures, to connet these to language processing, and to gain insight into online language processing (Conklin, Pellicer-Sánchez, and Carrol, 2018). The current study aims to investigate how EFL university students process verb agreement in the sentences with subject of complex noun phrases. Two types of subject noun phrases were investigated: syntactic process induced with single token and semantic process induced with multiple token. Overall, analyses of eye-movements revealed that the participants showed distinct processing patterns for single and multiple tokens.

Abstract processes and conflicts in place/transition systems

For one-safe Petri nets or condition/event-systems, a process as defined by Carl Adam Petri provides a notion of a run of a system where causal dependencies are reflected in terms of a partial order. Goltz and Reisig have generalised this concept for nets where places carry multiple tokens, by distinguishing tokens according to their causal history. However, this so-called individual token interpretation is often considered too detailed. Here we identify a subclass of Petri nets, called structural conflict nets, where no interplay between conflict and concurrency due to token multiplicity occurs. For this subclass, we define abstract processes as equivalence classes of Goltz-Reisig processes. We justify this approach by showing that there is a largest abstract process if and only if the underlying net is conflict-free with respect to a canonical notion of conflict.

The Adoption of Blockchain-based Decentralized Exchanges: A Market Microstructure Analysis of the Automated Market Maker

We investigate the market microstructure of Automated Market Makers (AMMs), the most prominent type of blockchain-based decentralized exchanges. We show that the order execution mechanism yields token value loss for liquidity providers if token exchange rates are volatile. AMMs are adopted only if their token pairs are highly correlated, or of high personal use for investors. A pricing curve with higher curvature makes the arbitrage problem less severe but also decreases investors’ surplus. Pooling multiple tokens exacerbates the arbitrage problem. We provide statistical support for our main model implications using transaction-level data of AMMs.

Barter Machine: An Autonomous, Distributed Barter Exchange on the Ethereum Blockchain

Direct bartering allows for the trading of assets with or without the use of money. In this paper, we introduce a smart contract written in the Solidity language for the Ethereum blockchain that implements a distributed and autonomous direct barter exchange operated by crowds. Since ERC20 smart contract tokens are widely used for initial coin offerings (ICOs), our implementation provides autonomous bartering services for ERC20 tokens. The non-fungible ERC721 token, as well as Ethereum Name Service (ENS) name bartering, are also supported. Because finding a feasible bartering solution for bids involving multiple tokens is NP-hard in general, our exchange only provides escrow and swapping services. It assumes feasible solutions are submitted by bartering problem solvers from the crowd who run a solver locally on their machines. Bartering problem solvers are incentivized to submit solutions to the autonomous exchange by awarding them with excess tokens that are left over after granting the bidders the tokens they requested in their bids. Our system, called BarterMachine, can perhaps be used to facilitate a global barter economy. The BarterMachine prototype is available for testing on the Ethereum Ropsten network at https://bartermachine.github.io/bartermachine/ropsten/.

An Efficient High-Throughput LZ77-Based Decompressor in Reconfigurable Logic

To best leverage high-bandwidth storage and network technologies requires an improvement in the speed at which we can decompress data. We present a “refine and recycle” method applicable to LZ77-type decompressors that enables efficient high-bandwidth designs and present an implementation in reconfigurable logic. The method refines the write commands (for literal tokens) and read commands (for copy tokens) to a set of commands that target a single bank of block ram, and rather than performing all the dependency calculations saves logic by recycling (read) commands that return with an invalid result. A single “Snappy” decompressor implemented in reconfigurable logic leveraging this method is capable of processing multiple literal or copy tokens per cycle and achieves up to 7.2GB/s, which can keep pace with an NVMe device. The proposed method is about an order of magnitude faster and an order of magnitude more power efficient than a state-of-the-art single-core software implementation. The logic and block ram resources required by the decompressor are sufficiently low so that a set of these decompressors can be implemented on a single FPGA of reasonable size to keep up with the bandwidth provided by the most recent interface technologies.

Leveraging Multi-Token Entities in Document-Level Named Entity Recognition

Most state-of-the-art named entity recognition systems are designed to process each sentence within a document independently. These systems are easy to confuse entity types when the context information in a sentence is not sufficient enough. To utilize the context information within the whole document, most document-level work let neural networks on their own to learn the relation across sentences, which is not intuitive enough for us humans. In this paper, we divide entities to multi-token entities that contain multiple tokens and single-token entities that are composed of a single token. We propose that the context information of multi-token entities should be more reliable in document-level NER for news articles. We design a fusion attention mechanism which not only learns the semantic relevance between occurrences of the same token, but also focuses more on occurrences belonging to multi-tokens entities. To identify multi-token entities, we design an auxiliary task namely ‘Multi-token Entity Classification’ and perform this task simultaneously with document-level NER. This auxiliary task is simplified from NER and doesn't require extra annotation. Experimental results on the CoNLL-2003 dataset and OntoNotesnbm dataset show that our model outperforms state-of-the-art sentence-level and document-level NER methods.

The cover time of deterministic random walks for general transition probabilities

A deterministic random walk is a deterministic process analogous to a random walk. While there are some results on the cover time of the rotor-router model, which is a deterministic random walk corresponding to a simple random walk, nothing is known about the cover time of deterministic random walks emulating general transition probabilities. This paper is concerned with the shortest remaining time (SRT)-router model with multiple tokens, which is a deterministic process coping with general transition probabilities possibly containing irrational numbers. For the model, we give an upper bound of the cover time, which is the first result on the cover time of deterministic random walks for general transition probabilities. Our upper bound also improves the existing bounds for the rotor-router model in some cases.

Development of Text-Based Algorithm for Opioid Overdose Identification in EMS Data

Development of Text-Based Algorithm for Opioid Overdose Identification in EMS Data

Barter Machine: Defier of the Money

Direct bartering allows us to do trading of assets with or without the use of money. We contribute a smart contract written in Solidity language for the Ethereum blockchain that implements a distributed and autonomous direct barter exchange operated by crowds. Since ERC20 smart contract tokens are widely used for initial coin offerings (ICOs), our implementation provides autonomous bartering services for ERC20 tokens. Non-fungible ERC721 token as well as Ethereum Name Service (ENS) name bartering are also supported. Finding feasible bartering solution for bids involving multiple tokens is NP-hard in general. Our exchange just provides escrow and swapping services. It assumes feasible solutions are submit ted by bartering problem solvers from the crowd who run a solver locally on their machines. Bartering problem solvers are incentivized for submitting solutions to the autonomous exchange by awarding them with excess tokens that are left over after granting the bidders the tokens they requested in their bids. Our system, called BarterMachine, can perhaps be used to facilitate a global barter economy. BarterMachine prototype is available for testing on the Ethereum Ropsten network.