arXiv subjects

Pidong Wang

Publications and source records attributed to Pidong Wang.

8 recordsLinked to original sources

arXiv2025

MetricX-25 and GemSpanEval: Google Translate Submissions to the WMT25 Evaluation Shared Task

In this paper, we present our submissions to the unified WMT25 Translation Evaluation Shared Task. For the Quality Score Prediction subtask, we create a new generation of MetricX with improvements in the input format and the training protocol, while for the Error Span Detection subtask we develop a new model, GemSpanEval, trained to predict error spans along with their severities and categories. Both systems are based on the state-of-the-art multilingual open-weights model Gemma 3, fine-tuned on publicly available WMT data. We demonstrate that MetricX-25, adapting Gemma 3 to an encoder-only architecture with a regression head on top, can be trained to effectively predict both MQM and ESA quality scores, and significantly outperforms its predecessor. Our decoder-only GemSpanEval model, on the other hand, we show to be competitive in error span detection with xCOMET, a strong encoder-only sequence-tagging baseline. With error span detection formulated as a generative task, we instruct the model to also output the context for each predicted error span, thus ensuring that error spans are identified unambiguously.

cs.CL↗

arXiv2022

Building Machine Translation Systems for the Next Thousand Languages

In this paper we share findings from our effort to build practical machine translation (MT) systems capable of translating across over one thousand languages. We describe results in three research domains: (i) Building clean, web-mined datasets for 1500+ languages by leveraging semi-supervised pre-training for language identification and developing data-driven filtering techniques; (ii) Developing practical MT models for under-served languages by leveraging massively multilingual models trained with supervised parallel data for over 100 high-resource languages and monolingual datasets for an additional 1000+ languages; and (iii) Studying the limitations of evaluation metrics for these languages and conducting qualitative analysis of the outputs from our MT models, highlighting several frequent error modes of these types of models. We hope that our work provides useful insights to practitioners working towards building MT systems for currently understudied languages, and highlights research directions that can complement the weaknesses of massively multilingual models in data-sparse settings.

cs.CL↗

arXiv2022

The Implicit Length Bias of Label Smoothing on Beam Search Decoding

Label smoothing is ubiquitously applied in Neural Machine Translation (NMT) training. While label smoothing offers a desired regularization effect during model training, in this paper we demonstrate that it nevertheless introduces length biases in the beam search decoding procedure. Our analysis shows that label smoothing implicitly applies a length penalty term to output sequence, causing a bias towards shorter translations. We also show that for a model fully optimized with label smoothing, translation length is implicitly upper bounded by a fixed constant independent of input. We verify our theory by applying a simple rectification function at inference time to restore the unbiased distributions from the label-smoothed model predictions. This rectification method led to consistent quality improvements on WMT English-German, English-French, English-Czech and English-Chinese tasks, up to +0.3 BLEU at beam size 4 and +2.8 BLEU at beam size 200.

cs.CL↗

arXiv2022

Multilingual Mix: Example Interpolation Improves Multilingual Neural Machine Translation

Multilingual neural machine translation models are trained to maximize the likelihood of a mix of examples drawn from multiple language pairs. The dominant inductive bias applied to these models is a shared vocabulary and a shared set of parameters across languages; the inputs and labels corresponding to examples drawn from different language pairs might still reside in distinct sub-spaces. In this paper, we introduce multilingual crossover encoder-decoder (mXEncDec) to fuse language pairs at an instance level. Our approach interpolates instances from different language pairs into joint `crossover examples' in order to encourage sharing input and output spaces across languages. To ensure better fusion of examples in multilingual settings, we propose several techniques to improve example interpolation across dissimilar languages under heavy data imbalance. Experiments on a large-scale WMT multilingual dataset demonstrate that our approach significantly improves quality on English-to-Many, Many-to-English and zero-shot translation tasks (from +0.5 BLEU up to +5.5 BLEU points). Results on code-switching sets demonstrate the capability of our approach to improve model generalization to out-of-distribution multilingual examples. We also conduct qualitative and quantitative representation comparisons to analyze the advantages of our approach at the representation level.

cs.CL↗

arXiv2021

Authentication of optical physical unclonable functions based on single-pixel detection

Physical unclonable function (PUF) has been proposed as a promising and trustworthy solution to a variety of cryptographic applications. Here we propose a non-imaging based authentication scheme for optical PUFs materialized by random scattering media, in which the characteristic fingerprints of optical PUFs are extracted from stochastical fluctuations of the scattered light intensity with respect to laser challenges which are detected by a single-pixel detector. The randomness, uniqueness, unpredictability, and robustness of the extracted fingerprints are validated to be qualified for real authentication applications. By increasing the key length and improving the signal to noise ratio, the false accept rate of a fake PUF can be dramatically lowered to the order of 10^-28. In comparison to the conventional laser-speckle-imaging based authentication with unique identity information obtained from textures of laser speckle patterns, this non-imaging scheme can be implemented at small speckle size bellowing the Nyquist--Shannon sampling criterion of the commonly used CCD or CMOS cameras, offering benefits in system miniaturization and immunity against reverse engineering attacks simultaneously.

cs.CR↗

arXiv2021

Bionic Optical Physical Unclonable Functions for Authentication and Encryption

Information security is of great importance for modern society with all things connected. Physical unclonable function (PUF) as a promising hardware primitive has been intensively studied for information security. However, the widely investigated silicon PUF with low entropy is vulnerable to various attacks. Herein, we introduce a concept of bionic optical PUFs inspired from unique biological architectures, and fabricate four types of bionic PUFs by molding the surface micro-nano structures of natural plant tissues with a simple, low-cost, green and environmentally friendly manufacturing process. The laser speckle responses of all bionic PUFs are statistically demonstrated to be random, unique, unpredictable and robust enough for cryptographic applications, indicating the broad applicability of bionic PUFs. On this ground, the feasibility of implementing bionic PUFs as cryptographic primitives in entity authentication and encrypted communication is experimentally validated, which shows its promising potential in the application of future information security.

cs.CR↗

arXiv2021

Fast random number generator based on optical physical unclonable functions

We propose an approach for fast random number generation based on homemade optical physical unclonable functions (PUFs). The optical PUF is illuminated with input laser wavefront of continuous modulation to obtain different speckle patterns. Random numbers are fully extracted from speckle patterns through a simple post-processing algorithm. Our proof-of-principle experiment achieves total random number generation rate of 0.96 Gbit/s with verified randomness, which is far faster than previous optical-PUF-based schemes. Our results demonstrate that the presented random number generator (RNG) proposal has great potential to achieve ultrafast random number generation rate up to several hundreds of Gbit/s.

cs.CR↗

arXiv2014

Plasmonic Tamm states: second enhancement of light inside the plasmonic waveguide

A type of Tamm states inside metal-insulator-metal (MIM) waveguides is proposed. An impedance based transfer matrix method is adopted to study and optimize it. With the participation of the plasmonic Tamm states, fields could be enhanced twice: the ffirst is due to the coupling between a normal waveguide and a nanoscaled plasmonic waveguide and the second is due to the strong localization and field enhancement of Tamm states. As shown in our 2D coupling configuration, |E|^2 is enhanced up to 1050 times when 1550 nm light is coupled from an 300 nm Si slab waveguide into an 40 nm MIM waveguide.

physics.optics↗