Anchorless Diversification for Parallel LLM Ideation
For practitioners using LLMs for open-ended ideation, this work provides practical, cost-effective diversification strategies without requiring seed ideas.
This paper investigates anchorless methods for diversifying LLM-generated candidate pools in creative ideation tasks, finding that semantic direction stratification achieves the best diversity-quality-compute trade-off, outperforming anchored baselines when accounting for full token usage.
LLMs are increasingly used to generate candidate-idea pools for creative tasks where broad exploration is valuable. Parallel inference can be attractive in this setting when it broadens the pool while retaining quality and cost efficiency. We study inference-time controls for candidate-pool diversification, asking whether anchorless methods can rival methods that depend on observed seed ideas. Across three creative task families, we compare independent generation and semantic direction stratification with self-, peer-, and representative-anchor baselines, under neutral and population-referential divergent instructions. Population-referential divergence is a strong low-cost baseline, increasing semantic diversity while preserving quality proxies. Semantic direction stratification is stronger: a single planning call organizes generations across broad semantic directions, yielding the best diversity--quality--compute frontier. Anchored regeneration can be strong in final-pool diversity, but its advantage shrinks under full-pipeline token accounting. These results establish practical anchorless baselines for open-ended LLM ideation.