paper-with-me

Papers

Pretraining Text Encoders with Adversarial Mixture of Training Signal Generators

2022-04-07 · ICLR 2022 4 · Yu Meng, Chenyan Xiong, Payal Bajaj, Saurabh Tiwary, Paul Bennett, Jiawei Han, Xia Song

We present a new framework AMOS that pretrains text encoders with an Adversarial learning curriculum via a Mixture Of Signals from multiple auxiliary generators. Following ELECTRA-style pretraining, the main encoder is trained as a discriminator to detect replaced tokens generated by auxiliary masked language models (MLMs). Different from ELECTRA which trains one MLM as the generator, we jointly train multiple MLMs of different sizes to provide training signals at various levels of difficulty. To push the discriminator to learn better with challenging replaced tokens, we learn mixture weights over the auxiliary MLMs' outputs to maximize the discriminator loss by backpropagating the gradient from the discriminator via Gumbel-Softmax. For better pretraining efficiency, we propose a way to assemble multiple MLMs into one unified auxiliary model. AMOS outperforms ELECTRA and recent state-of-the-art pretrained models by about 1 point on the GLUE benchmark for BERT base-sized models.

📄 PDF Abstract BibTeX arXiv:2204.03243

Code (1)

microsoft/amos 공식 구현 pytorch

Methods 이 논문이 사용한 방법론

Refunds@Expedia|||How do I get a full refund from Expedia? “How do I get a full refund from Expedia? How do I get a full refund from Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Quick Help &…
Multi-Head Attention 설명 없음
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Attention Dropout Attention Dropout is a type of dropout used in attention-based architectures, where elements are randomly dropped out of the…
WordPiece 설명 없음
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…

Similar Papers 제목 키워드 기반

Improving Adversarial Robustness of Masked Autoencoders via Test-time Frequency-domain Prompting

2023-08-20 · ICCV 2023 1 · Qidong Huang, Xiaoyi Dong, Dongdong Chen, Yinpeng Chen 외

In this paper, we investigate the adversarial robustness of vision transformers that are equipped with BERT pretraining (e.g., BEiT, MAE). A surprising observation is that MAE has significantly worse adversarial robustne…

Adversarial Robustness

ShapleyLaw: A Game-Theoretic Approach to Multilingual Scaling Laws

2026-03-18 · Xuyang Cao, Qianying Liu, Chuan Xiao, Yusuke Oda 외 arxiv

In multilingual pretraining, the test loss of a pretrained model is heavily influenced by the proportion of each language in the pretraining data, namely the \textit{language mixture ratios}. Multilingual scaling laws ca…

Cross-Lingual Transfer

LLMSurgeon: Diagnosing Data Mixture of Large Language Models

2026-05-28 · Yaxin Luo, Jiacheng Cui, Xiaohan Zhao, Xinyi Shang 외 arxiv

The pretraining data mixture of Large Language Models (LLMs) constitutes their "digital DNA", shaping model behaviors, capabilities, and failure modes. Yet this composition is rarely disclosed, making post-hoc auditing o…

Double Backpropagation for Training Autoencoders against Adversarial Attack

2020-03-04 · Chengjin Sun, Sizhe Chen, Xiaolin Huang

Deep learning, as widely known, is vulnerable to adversarial samples. This paper focuses on the adversarial attack on autoencoders. Safety of the autoencoders (AEs) is important because they are widely used as a compress…

Adversarial AttackRobust classification

Pretraining Data Mixtures Enable Narrow Model Selection Capabilities in Transformer Models

2023-11-01 · Steve Yadlowsky, Lyric Doshi, Nilesh Tripuraneni

Transformer models, notably large language models (LLMs), have the remarkable ability to perform in-context learning (ICL) -- to perform new tasks when prompted with unseen input-output examples without any explicit mode…

In-Context LearningModel Selection