Re3: Generating Longer Stories With Recursive Reprompting and Revision

“Of course, recent years have also witnessed a dramatic rise in the capabilities of general-purpose (non-finetuned) large pretrained language models. Of particular note are their strong zero-shot capabilities, especially when given clever prompts (Brown et al., 2020; Kojima et al., 2022). Yet despite recent improvements, even the best models to date may still struggle with complex long-form generation, such as in our story generation task (Section 4).

In contrast, human writers successfully navigate the myriad challenges of long-form generation on a regular basis. We observe that a human writer does not simply write a long document in one shot. Rather, he or she may (a) create a detailed plan, then (b) draft each next passage of the document according to that plan. He or she may then revise by (c) rewriting passages entirely, and/or (d) postediting for finer details.

Motivated by this observation, we propose the Recursive Reprompting and Revision framework (Re3, Figure 1) to generate longer stories. While based on the human writing process, Re3 is a fully automatic system with no human intervention, unlike prior approaches which model the human writing process with a human in the loop (Goldfarb- Tarrant et al., 2019; Coenen et al., 2021; Lee et al., 2022). First, (a) Re3’s Plan module generates a plan by prompting GPT3 (Brown et al., 2020) to augment a given premise with a setting, characters, and outline. (b) Re3’s Draft module then generates each next story continuation by recursively reprompting GPT3 using a strategically crafted prompt, in a procedure which can be viewed as a generalization of chain-of-thought prompting (Kojima et al., 2022). Specifically, our prompt is dynamically reconstructed at each step by selectively manifesting contextually relevant information from the initial plan—itself generated by prompting—and the story thus far. We then divide the revision process into (c) a Rewrite module which emulates a full rewrite by reranking alternate continuations, and (d) an Edit module which makes smaller local edits to improve factual consistency with previous passages.

As an additional contribution, our Plan and Draft modules are fully zero-shot rather than trained on existing story datasets. Thus not only does Re3 generate stories an order of magnitude longer than those of prior work, but it is not limited to any particular training domain.”