Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “tokenization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3

Queueing models for token and slotted ring networks

Currently the end-to-end delay characteristics of very high speed local area networks are not well understood. The transmission speed of computer networks is increasing, and local area networks especially are finding increasing use in real time systems. Ring networks operation is generally well understood for both token rings and slotted rings. There is, however, a severe lack of queueing models for high layer operation. There are several factors which contribute to the processing delay of a packet, as opposed to the transmission delay, e.g., packet priority, its length, the user load, the processor load, the use of priority preemption, the use of preemption at packet reception, the number of processors, the number of protocol processing layers, the speed of each processor, and queue length limitations. Currently existing medium access queueing models are extended by adding modeling techniques which will handle exhaustive limited service both with and without priority traffic, and modeling capabilities are extended into the upper layers of the OSI model. Some of the model are parameterized solution methods, since it is shown that certain models do not exist as parameterized solutions, but rather as solution methods.

Peden, Jeffery H.↗

Did You Win the GPU Cloud Lottery? Benchmarking from TFLOPS to Tokens/$

Cloud GPUs are commonly assumed to deliver consistent performance for a given GPU model. This assumption does not always hold: cloud providers employ diverse system configurations and virtualization mechanisms, and GPUs themselves exhibit non-negligible manufacturing variability (the silicon lottery). In this work, we present a large-scale measurement study of GPU performance variability across 11 cloud providers, covering over 3,500 physical GPUs and 6,800 benchmark runs. Our hierarchical analysis shows that while execution-level variation stays below 9%, performance varies by up to 38% across devices and providers for the same GPU model. Regression analysis indicates that driver- and OS-related software factors contribute less than 1% of the variance; instead, silicon lottery effects dominate observed performance variation, and cloud providers further amplify them through persistent, systematic second-order effects.

Slynko, Platon [Silicon Data, New York, USA] (ORCI↗

Security Evaluation of Smart Cards and Secure Tokens: Benefits and Drawbacks for Reducing Supply Chain Risks of Nuclear Power Plants

The supply chain attack pathway is being increasingly used by adversaries to bypass security controls and gain unauthorized access to sensitive networks and equipment (e.g., Critical Digital Assets). Cyber-attacks targeting supply chain generally aim to compromise the environments, products, or services of vendors and suppliers to inject, add, or substitute authentic software and hardware with malicious elements. These malicious elements are deemed to be authentic as they arise from the vendor or supplier (i.e., the supply chain). This research aims to leverage findings and assumptions made from the previous report to determine the security benefits and drawbacks of a smart card- based hardware root of trust. Smart cards can provide devices inside Nuclear Power Plants (NPP) with a secure environment to store keys in and perform sensitive operations such as digital signature generation. These abilities can be leveraged to increase supply chain cybersecurity by autonomously providing NPP Licensees with reports on device integrity, authenticity and measurements of executable and non-executable data.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

An image compression technique for use on token ring networks

A low complexity technique for compression of images for transmission over local area networks is presented. The technique uses the synchronous traffic as a side channel for improving the performance of an adaptive differential pulse code modulation (ADPCM) based coder.

Gorjala, B.↗

Name Normalizer

Our name normalizer and enrichment micro-service address the shortcomings of standard text normalization. Initially the text input is split into tokens. This process takes into account some conventions of formatting. Any dashes present between two tokens preserves the relationship of those two tokens. Conversely a comma between two tokens ensures that the separation between the two tokens is maintained. The order of tokens is also preserved. Standard text normalization (conversion to lowercase, trimming extra white space, and canonicalization) is then applied to the tokens. Once normalized every unique and sequence of tokens is given confidence values by comparing the normalized values to publicly available data.

Bleeker, Amelia↗

Measuring the quality of a quantum reference frame: The relative entropy of frameness

In the absence of a reference frame for transformations associated with group G, any quantum state that is noninvariant under the action of G may serve as a token of the missing reference frame. We here present a measure of the quality of such a token: the relative entropy of frameness. This is defined as the relative entropy distance between the state of interest and the nearest G-invariant state. Unlike the relative entropy of entanglement, this quantity is straightforward to calculate, and we find it to be precisely equal to the G-asymmetry, a measure of frameness introduced by Vaccaro et al. It is shown to provide an upper bound on the mutual information between the group element encoded into the token and the group element that may be extracted from it by measurement. In this sense, it quantifies the extent to which the token successfully simulates a full reference frame. We also show that despite a suggestive analogy from entanglement theory, the regularized relative entropy of frameness is zero and therefore does not quantify the rate of interconversion between the token and some standard form of quantum reference frame. Finally, we show how these investigations yield an approach to bounding the relative entropy of entanglement.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

MTRE

Multi-Token Reliability Estimation (MTRE) is a lightweight, white-box hallucination detector for vision-language models. Instead of using only the first output token, MTRE aggregates logits from the first ~10 tokens and feeds them to a small attention-based reliability head; per-token scores are combined via a sequential log-likelihood-ratio test with early-stopping, and an MTRE-t variant calibrates thresholds via cross-fitting. MTRE reports average gains of +9.4% Accuracy and +14.8% AUROC over common baselines across MAD-Bench, MM-SafetyBench, MathVista, and arithmetic/counting tasks, while adding ~4.3M params and ~1% inference overhead (~26 MB VRAM, ~0.94 ms per detection). Key limitation: requires access to early token logits and is evaluated on a handful of open-source 7B VLMs.

Bhattarai, Manish [Los Alamos National Labs]↗

MATEY: multiscale adaptive transformer models for spatiotemporal physical systems

Accurate representation of the multiscale features in spatiotemporal physical systems using vision transformer architectures requires extremely long, computationally prohibitive token sequences. To address this issue, we propose two novel adaptive tokenization schemes that dynamically adjust patch sizes based on local features: one ensures convergent behavior to uniform patch refinement, while the other offers better computational efficiency. Moreover, we present a set of spatiotemporal attention schemes, where the temporal or axial spatial dimensions are decoupled, to evaluate their baseline computational and data efficiencies and to determine whether adaptive tokenization can improve this performance. We assess the performance of the proposed multiscale adaptive model, MATEY, in a sequence of experiments. Compared to a full spatiotemporal attention scheme or a scheme that decouples only the temporal dimension, we find that fully decoupled axial attention is less efficient and expressive, requiring more training time and model parameters to achieve the same accuracy. The experiments on the adaptive tokenization schemes show that, compared to a uniformly refined model, the proposed schemes achieve comparable or improved accuracy at a much lower cost in the tested two-dimensional settings. While the asymptotic analysis suggests the potential for favorable scaling, empirical validation at substantially longer sequence lengths remains to be performed in future work. Finally, we demonstrate in two fine-tuning tasks featuring different physics that models pretrained on PDEBench data outperform the ones trained from scratch, especially in the low data regime with frozen attention.

adaptive tokenization↗

Performance analysis of FDDI

The Fiber Distributed Data Interface (FDDI) is an imerging ANSI and ISO standard for a 100 megabit per second fiber optic token ring. The performance of the FDDI media access control protocol is analyzed using a simulation developed at NASA Ames. Both analyses using standard measures of performance (including average delay for asynchronous traffic, channel utilization, and transmission queue length) and analyses of characteristics of ring behavior which can be attributed to constraints imposed by the timed token protocol on token holding time (including bounded token rotation time, support for synchronous traffic, and fairness of channel access for nodes transmitting asynchronous traffic) are included.

Johnson, Marjory J.↗

Fiber optic voice/data network

An asynchronous, high-speed, fiber optic local area network originally developed for tactical environments with additional benefits for other environments such as spacecraft, and the like. The network supports ordinary data packet traffic simultaneously with synchronous T1 voice traffic over a common token ring channel; however, the techniques and apparatus of this invention can be applied to any deterministic class of packet data networks, including multitier backbones, that must transport stream data (e.g., video, SAR, sensors) as well as data. A voice interface module parses, buffers, and resynchronizes the voice data to the packet network employing elastic buffers on both the sending and receiving ends. Voice call setup and switching functions are performed external to the network with ordinary PABX equipment. Clock information is passed across network boundaries in a token passing ring by preceeding the token with an idle period of non-transmission which allows the token to be used to re-establish a clock synchronized to the data. Provision is made to monitor and compensate the elastic receiving buffers so as to prevent them from overflowing or going empty.

Bergman, Larry A.↗

Debiasing Watermarks for Large Language Models via Maximal Coupling

Watermarking language models is essential for distinguishing between human and machine-generated text and thus maintaining the integrity and trustworthiness of digital communication. Here, we present a novel green/red list watermarking approach that partitions the token set into “green” and “red” lists, subtly increasing the generation probability for green tokens. To correct token distribution bias, our method employs maximal coupling, using a uniform coin flip to decide whether to apply bias correction, with the result embedded as a pseudorandom watermark signal. Theoretical analysis confirms this approach’s unbiased nature and robust detection capabilities. Experimental results show that it outperforms prior techniques by preserving text quality while maintaining high detectability, and it demonstrates resilience to targeted modifications aimed at improving text quality. This research provides a promising watermarking solution for language models, balancing effective detection with minimal impact on text quality.

97 MATHEMATICS AND COMPUTING↗

Position-Enhanced Gradient Attack (PEGA) on Medical Language Models

Federated Learning (FL) enables collaborative training of language models on sensitive clinical notes without sharing the data. However, this paradigm is vulnerable to gradient inversion attacks that can reconstruct private data from shared gradients. We find that state-of-the-art attacks are less effective in the medical domain, failing to overcome the unique challenges posed by its specialized vocabulary and unstructured format. To address this, we introduce the Position-Enhanced Gradient Attack (PEGA), a novel attack that makes gradients position-aware by optimizing token and position embeddings simultaneously. PEGA employs two key innovations: a periodic sorting of positional embeddings to resolve token order ambiguity and a late-stage embedding replacement strategy to correct hard-to-recover critical tokens. To evaluate the leakage of sensitive data more directly, we also propose the Unified PHI-Recall (UPHI), a new metric measuring the recovery of Protected Health Information. Experiments on the MIMIC-III dataset show that PEGA significantly outperforms leading attacks like TAG and LAMP, particularly in its ability to reconstruct identifiable patient information, exposing a more severe and nuanced privacy risk in federated medical NLP.

Xu, Nuo [University of Minnesota]↗

XMark: Reliable Multi-Bit Watermarking for LLM-Generated Texts

Multi-bit watermarking has emerged as a promising solution for embedding imperceptible binary messages into Large Language Model (LLM)-generated text, enabling reliable attribution and tracing of malicious usage of LLMs. Despite recent progress, existing methods still face key limitations: some become computationally infeasible for large messages, while others suffer from a poor trade-off between text quality and decoding accuracy. Moreover, the decoding accuracy of existing methods drops significantly when the number of tokens in the generated text is limited, a condition that frequently arises in practical usage. To address these challenges, we propose XMark, a novel method for encoding and decoding binary messages in LLM-generated texts. The unique design of XMark’s encoder produces a less distorted logit distribution for watermarked token generation, preserving text quality, and also enables its tailored decoder to reliably recover the encoded message with limited tokens. Extensive experiments across diverse downstream tasks show that XMark significantly improves decoding accuracy while preserving the quality of watermarked text, outperforming prior methods. The code will be made publicly available upon acceptance.

Xu, Jiahao [University of Nevada, Reno]↗

An extended vector space model for information retrieval with generalized similarity measures: theory and applications

We present an Extended Vector Space Model (EVSM) for information retrieval, endowed with a new set of similarity functions. Our model considers records as multisets of tokens. A token weight function maps records into a real vectors. Using this vector representation we define a p -norm of a record and pairwise conjunction and disjunction operations on records. These oper ations prompt consistent extensions of published set-based similarity functions and yield new $\ell$ p distance-based similarities. We demonstrate that some well-known similarities form a subset of the new functions resulting from particular choices of token weights and p -values.

Paskaleva, Biliana Stefanova↗