P${}^3$LM: Probabilistically Permuted Prophet Language Modeling For Generative Pre-training

Bao Junwei, Wang Yifan, Ying Jiangyong, Gong Yeyun, Zhao Jing, Wu Youzheng, He Xiaodong. Arxiv 2022

[Paper]    
Applications Attention Mechanism GPT Language Modeling Model Architecture Pretraining Methods Training Techniques Transformer

Conventional autoregressive left-to-right (L2R) sequence generation faces two issues during decoding: limited to unidirectional target sequence modeling, and constrained on strong local dependencies. To address the aforementioned problem, we propose P${}^3$LM, a probabilistically permuted prophet language model, which strengthens the modeling of bidirectional information and long token dependencies for sequence generation. Specifically, P${}^3$LM learns to generate tokens in permuted order upon an order-aware transformer decoder, as well as to generate the corresponding future $N$ tokens with a multi-stream attention mechanism. Extensive experiments are conducted on the GLGE benchmark, which includes four datasets for summarization, two for question generation, one for conversational question answering, and one for dialog response generation, where P${}^3$LM achieves state-of-the-art results compared with strong publicly available generative pre-training methods.

Similar Work

• Retrieval-pretrained Transformer: Long-range Language Modeling With Self-retrieval
• REPLUG: Retrieval-augmented Black-box Language Models
• Prophet: Prompting Large Language Models With Complementary Answer Heuristics For Knowledge-based Visual Question Answering
• Probabilistically Masked Language Model Capable Of Autoregressive Generation In Arbitrary Word Order