Inducing Sustained Creativity and Diversity in Large Language Models

2026-03-19

AI Summary

This work addresses the tendency of current large language models to produce repetitive or conventional outputs during exploratory search, hindering sustained generation of diverse and creative content. The authors propose a novel decoding mechanism that operates without access to internal model representations, instead leveraging external feedback to dynamically steer the generation trajectory. This approach effectively enhances the model’s capacity to combine both orthodox and unorthodox knowledge, moving beyond traditional decoding strategies narrowly optimized for correctness. Notably, it achieves prolonged maintenance of high output diversity and novelty without any architectural modifications to the underlying model. As a result, the method significantly improves users’ efficiency in discovering satisfactory solutions within complex search spaces and accommodates personalized exploration needs.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Learning to Search
• Natural Language Processing: (Large) Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

We address a not-widely-recognized subset of exploratory search, where a user sets out on a typically long "search quest" for the perfect wedding dress, overlooked research topic, killer company idea, etc. The first few outputs of current large language models (LLMs) may be helpful but only as a start, since the quest requires learning the search space and evaluating many diverse and creative alternatives along the way. Although LLMs encode an impressive fraction of the world's knowledge, common decoding methods are narrowly optimized for prompts with correct answers and thus return mostly homogeneous and conventional results. Other approaches, including those designed to increase diversity across a small set of answers, start to repeat themselves long before search quest users learn enough to make final choices, or offer a uniform type of "creativity" to every user asking similar questions. We develop a novel, easy-to-implement decoding scheme that induces sustained creativity and diversity in LLMs, producing as many conceptually unique results as desired, even without access to the inner workings of an LLM's vector space. The algorithm unlocks an LLM's vast knowledge, both orthodox and heterodox, well beyond modal decoding paths. With this approach, search quest users can more quickly explore the search space and find satisfying answers.

Problem

• exploratory search
• creativity
• diversity
• large language models
• search quest

Innovation

• sustained creativity
• diversity
• decoding scheme
• exploratory search
• large language models