Back to AI News
arxiv

AutoManifold Uses Multi-Agent LLMs to Design Manifold Embedding Algorithms

AutoManifold is an agentic pipeline that composes new manifold embedding algorithms from constrained affinity, cost, and optimisation primitives using multi-agent LLM orchestration. It conditions each design stage on user-specified structural preservation preferences and refines configurations through an LLM-guided, metric-grounded iterative loop. The authors compare generated algorithms with t-SNE, UMAP, and PaCMAP on real-world datasets under identical evaluation infrastructure. The paper reports that priority-conditioned synthesis can produce new algorithms that steer toward user-specified properties without compromising intrinsic data structure, and the code is publicly available.

AI Brief

It demonstrates a concrete LLM-driven approach to autonomously generating and evaluating data-visualisation algorithms rather than only tuning existing ones.

arXiv:2611.13710v1 Announce Type: new Abstract: Manifold embedding for data visualisation is dominated by a small set of hand-designed state-of-the-art (SOTA) algorithms. Their design targets are fixed at development time, resulting in inflexible algorithmic behaviour that cannot be readily steered towards user-specified structural priorities. Hand-designing new embedding algorithms to address user priority requires highly specialised expertise and is time consuming. To address autonomous algorithm design tailored to user preference, we introduce an agentic algorithm generation pipeline AutoManifold. It composes new manifold embedding algorithms from a constrained vocabulary of affinity, cost, and optimisation primitives, supported by multi-agent large-language-model (LLM) orchestration. AutoManifold conditions every stage of its design on user-specified structural preservation preferences, and iteratively refines the algorithm configuration through an LLM-guided, metric-grounded iterative loop. We compare the generated algorithms against the three strongest and most frequently used SOTA methods (t-SNE, UMAP, and PaCMAP) on real-world datasets spanning different difficulty regimes under identical evaluation infrastructure. The results demonstrate that LLM-driven, priority-conditioned algorithm synthesis can move beyond hyperparameter tuning, producing genuinely new algorithms that outperform traditional hand-designed methods in their ability to steer towards user-specified properties without compromising the intrinsic structure in the original data. The code of AutoManifold is available at https://github.com/buraksusam/AutoManifold_NeurIPS2026.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!