English

Text Editing as Imitation Game

Computation and Language 2022-10-25 v1

Abstract

Text editing, such as grammatical error correction, arises naturally from imperfect textual data. Recent works frame text editing as a multi-round sequence tagging task, where operations -- such as insertion and substitution -- are represented as a sequence of tags. While achieving good results, this encoding is limited in flexibility as all actions are bound to token-level tags. In this work, we reformulate text editing as an imitation game using behavioral cloning. Specifically, we convert conventional sequence-to-sequence data into state-to-action demonstrations, where the action space can be as flexible as needed. Instead of generating the actions one at a time, we introduce a dual decoders structure to parallel the decoding while retaining the dependencies between action tokens, coupled with trajectory augmentation to alleviate the distribution shift that imitation learning often suffers. In experiments on a suite of Arithmetic Equation benchmarks, our model consistently outperforms the autoregressive baselines in terms of performance, efficiency, and robustness. We hope our findings will shed light on future studies in reinforcement learning applying sequence-level action generation to natural language processing.

Keywords

Cite

@article{arxiv.2210.12276,
  title  = {Text Editing as Imitation Game},
  author = {Ning Shi and Bin Tang and Bo Yuan and Longtao Huang and Yewen Pu and Jie Fu and Zhouhan Lin},
  journal= {arXiv preprint arXiv:2210.12276},
  year   = {2022}
}

Comments

Accepted to Findings of EMNLP 2022

R2 v1 2026-06-28T04:13:36.062Z