# COD3S: Diverse Generation with Discrete Semantic Signatures

arXiv: https://arxiv.org/abs/2010.02882 · HTML: https://mlsanity.com/paper/2010.02882/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Oct 2020.

## Built on by (most cited first, 9 of 9)

- [Tailor: Generating and Perturbing Text with Semantic Controls](https://mlsanity.com/paper/2107.07150/) (2021)
- [What Comes Next? Evaluating Uncertainty in Neural Text Generators Against Human Production Variability](https://mlsanity.com/paper/2305.11707/) (2023)
- [Determinantal Beam Search](https://mlsanity.com/paper/2106.07400/) (2021)
- [ALERT: Adapting Language Models to Reasoning Tasks](https://mlsanity.com/paper/2212.08286/) (2022)
- [PCFG-based Natural Language Interface Improves Generalization for Controlled Text Generation](https://mlsanity.com/paper/2210.07431/) (2022)
- [A survey of diversity quantification in natural language processing: The why, what, where and how](https://mlsanity.com/paper/2507.20858/) (2025)
- [InFillmore: Frame-Guided Language Generation with Bidirectional Context](https://mlsanity.com/paper/2103.04941/) (2021)
- [Toward Diverse Precondition Generation](https://mlsanity.com/paper/2106.07117/) (2021)
- [Modeling Complex Event Scenarios via Simple Entity-focused Questions](https://mlsanity.com/paper/2302.07139/) (2023)

## Abstract (arXiv)

We present COD3S, a novel method for generating semantically diverse sentences using neural sequence-to-sequence (seq2seq) models. Conditioned on an input, seq2seq models typically produce semantically and syntactically homogeneous sets of sentences and thus perform poorly on one-to-many sequence generation tasks. Our two-stage approach improves output diversity by conditioning generation on locality-sensitive hash (LSH)-based semantic sentence codes whose Hamming distances highly correlate with human judgments of semantic textual similarity. Though it is generally applicable, we apply COD3S to causal generation, the task of predicting a proposition's plausible causes or effects. We demonstrate through automatic and human evaluation that responses produced using our method exhibit improved diversity without degrading task performance.