Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
3 results for “controllability”
StorySpark: Module-wise Evolutionary Search for Story Premise Generation
StorySpark is a new AI research method introduced on arXiv that uses evolutionary search over modular narrative components (e.g., background, persona, twist) to generate more original and high-quality story premises than existing LLM-based approaches.
Aug 14, 2026
Exploring More to Solve More: Boosting Diversity in Text Diffusion Models via Entropy-Based Guidance
Researchers propose a new training-free guidance method called SAKE for text diffusion models that uses entropy-based semantic analysis to improve the balance between output fidelity and diversity, with demonstrated gains on code and math generation tasks.
Aug 4, 2026
Behavioral Controllability of Agentic Models for Information Extraction: From Fixed Workflows to Reflective Agents
A new arXiv preprint investigates whether reflective LLM agents improve controllability and observable behavior over fixed workflows in scholarly dataset extraction, using process-level metrics rather than just accuracy.
Jul 20, 2026