aiqbench analysis
This research introduces DivLM, a post-training framework designed to enhance diversity in large language model-generated short stories across genre, tone, and style without sacrificing quality. By combining continued pre-training with reinforcement learning using a composite reward function, the method achieves over a 9% improvement in diversity metrics while maintaining instruction-following capabilities. The work addresses a critical limitation in current generative models where accuracy often comes at the expense of creative variation.
Improving Diversity in LLM Short Story Generation
arXiv:2610.06729v1 Announce Type: new Abstract: Large language models (LLMs) can generate accurate responses, but these are void of diversity. We attempt to address this for the task of creative short story generation. Drawing on established writing conventions and known LLM limitations, we target variation in genre, tone, style, and named entities. To promote diversity across these dimensions, we introduce DivLM, an LLM post-training framework consisting of two phases. First, we perform continued pre-training on a creative writing corpus and restore instruction-following capabilities using weight residuals. We then apply reinforcement learning with a custom, composite reward function that jointly maximizes diversity across the targeted narrative dimensions while maintaining response quality. Our empirical results on two LLM families show that DivLM increases diversity metrics by more than 9% on average compared to alternative approaches, while preserving instruction following, overall response quality, and similarity to human outputs.
Abstract:Large language models (LLMs) can generate accurate responses, but these are void of diversity. We attempt to address this for the task of creative short story generation. Drawing on established writing conventions and known LLM limitations, we target variation in genre, tone, style, and named entities. To promote diversity across these dimensions, we introduce DivLM, an LLM post-training framework consisting of two phases. First, we perform continued pre-training on a creative writing corpus and restore instruction-following capabilities using weight residuals. We then apply reinforcement learning with a custom, composite reward function that jointly maximizes diversity across the targeted narrative dimensions while maintaining response quality. Our empirical results on two LLM families show that DivLM increases diversity metrics by more than 9% on average compared to alternative approaches, while preserving instruction following, overall response quality, and similarity to human outputs.
Computation and Language (cs.CL)
arXiv:2610.06729 [cs.CL]
(or arXiv:2610.06729v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2610.06729 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 234 words
Related coverage
[AINews] Opus 5.5 is good at explainer videos
Latent Space · Sep 29, 2026
[AINews] The Future of Latent Space
Latent Space · Sep 25, 2026
Building and evaluating model diffing agents
AI Alignment Forum · Jun 12, 2026
[AINews] Xiaomi MiMo-V2.6-Pro 1T-A42B: the new top Open Weights model, trained for $3M
Latent Space · Sep 22, 2026