Don’t Add, don’t Miss: Effective Content Preserving Generation from Pre-Selected Text Spans

Aviv Slobodkin, Avi Caciularu, Eran Hirsch, Ido Dagan

Submission Type: Regular Long Paper

Submission Track: Natural Language Generation

Submission Track 2: Summarization

Keywords: NLG, Controlled Generation, RL, Controlled Decoding, Distillation

Abstract:

The recently introduced Controlled Text Reduction (CTR) task isolates the text generation step within typical summarization-style tasks. It does so by challenging models to generate coherent text conforming to pre-selected content within the input text (``highlights'').
This framing enables increased modularity in summarization-like tasks, allowing to couple a single CTR model with various content-selection setups and modules.
However, there are currently no reliable CTR models, while the performance of the existing baseline for the task is mediocre, falling short of practical utility.
Here, we address this gap by introducing a high-quality, open-source CTR model that tackles two prior key limitations: inadequate enforcement of the content-preservation constraint, and suboptimal silver training data.
Addressing these, we amplify the content-preservation constraint in both training, via RL, and inference, via a controlled decoding strategy.
Further, we substantially improve the silver training data quality via GPT-4 distillation.
Overall, pairing the distilled dataset with the highlight-adherence strategies yields marked gains over the current baseline, of up to 30 ROUGE-L points, providing a reliable CTR model for downstream use.

Submission Number: 3726