References & Citations
Computer Science > Computation and Language
Title: A Contrastive Framework for Neural Text Generation
(Submitted on 13 Feb 2022 (v1), revised 30 May 2022 (this version, v2), latest version 26 Sep 2022 (v3))
Abstract: Text generation is of great importance to many natural language processing applications. However, maximization-based decoding methods (e.g. beam search) of neural language models often lead to degenerate solutions -- the generated text is unnatural and contains undesirable repetitions. Existing approaches introduce stochasticity via sampling or modify training objectives to decrease probabilities of certain tokens (e.g., unlikelihood training). However, they often lead to solutions that lack coherence. In this work, we show that an underlying reason for model degeneration is the anisotropic distribution of token representations. We present a contrastive solution: (i) SimCTG, a contrastive training objective to calibrate the model's representation space, and (ii) a decoding method -- contrastive search -- to encourage diversity while maintaining coherence in the generated text. Extensive experiments and analyses on three benchmarks from two languages demonstrate that our proposed approach outperforms state-of-the-art text generation methods as evaluated by both human and automatic metrics.
Submission history
From: Yixuan Su [view email][v1] Sun, 13 Feb 2022 21:46:14 GMT (5592kb,D)
[v2] Mon, 30 May 2022 08:25:58 GMT (2809kb,D)
[v3] Mon, 26 Sep 2022 14:59:44 GMT (2854kb,D)
Link back to: arXiv, form interface, contact.