analiza aiqbench
This paper proposes a fine-tuning method for pretrained diffusion and flow-matching generators that optimizes the expected hypervolume of a batch of samples, so candidates are scored by how well they complement each other and not just individually. The authors show this objective can't be written as a fixed per-sample reward, and they work around it with an adaptive reward estimator and KL-regularized subproblems with stated convergence guarantees. The reported gains are on therapeutic-peptide and molecular design benchmarks, so readers should weigh the results against how those tasks are evaluated.
Multi-Objective Generative Optimization via Expected Hypervolume Fine-Tuning
arXiv:2611.05207v1 Announce Type: new Abstract: Scientific discovery requires plausible designs spanning useful trade-offs among competing properties: a molecule with strong binding affinity, for example, may be difficult to synthesize. Pretrained generative models, e.g., flow and diffusion generators, provide a foundation for proposing such designs, but steering them toward high-quality, complementary trade-offs remains challenging. Inference-time search incurs high per-sample costs, while fine-tuning typically relies on fixed per-candidate rewards that do not capture how designs complement one another. We introduce Hypervolume Fine-Tuning (HVFT), which adapts a single pretrained generator to maximize the expected hypervolume of its samples while controlling departure from the pretrained distribution. Expected hypervolume jointly rewards candidate quality and collective trade-off coverage. Nonetheless, we show that it cannot be expressed as the expectation of a fixed per-candidate reward and therefore cannot be optimized directly with standard techniques. A key observation of this work is that its first variation corresponds to the expected marginal hypervolume contribution of a candidate, which yields a tractable adaptive reward via an unbiased estimator based on model samples. HVFT uses this reward to decompose the objective into a sequence of KL-proximal scalar-reward subproblems. We establish concavity, relative smoothness, and convergence guarantees for the resulting distributional procedure. Across therapeutic-peptide and molecular design tasks spanning discrete diffusion and continuous flow-matching models, HVFT achieves state-of-the-art expected hypervolume and trade-off coverage.
Abstract:Scientific discovery requires plausible designs spanning useful trade-offs among competing properties: a molecule with strong binding affinity, for example, may be difficult to synthesize. Pretrained generative models, e.g., flow and diffusion generators, provide a foundation for proposing such designs, but steering them toward high-quality, complementary trade-offs remains challenging. Inference-time search incurs high per-sample costs, while fine-tuning typically relies on fixed per-candidate rewards that do not capture how designs complement one another. We introduce Hypervolume Fine-Tuning (HVFT), which adapts a single pretrained generator to maximize the expected hypervolume of its samples while controlling departure from the pretrained distribution. Expected hypervolume jointly rewards candidate quality and collective trade-off coverage. Nonetheless, we show that it cannot be expressed as the expectation of a fixed per-candidate reward and therefore cannot be optimized directly with standard techniques. A key observation of this work is that its first variation corresponds to the expected marginal hypervolume contribution of a candidate, which yields a tractable adaptive reward via an unbiased estimator based on model samples. HVFT uses this reward to decompose the objective into a sequence of KL-proximal scalar-reward subproblems. We establish concavity, relative smoothness, and convergence guarantees for the resulting distributional procedure. Across therapeutic-peptide and molecular design tasks spanning discrete diffusion and continuous flow-matching models, HVFT achieves state-of-the-art expected hypervolume and trade-off coverage.
Machine Learning (cs.LG)
arXiv:2611.05207 [cs.LG]
(or arXiv:2611.05207v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2611.05207 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.
Text integral sub licența deschisă a publicației · 310 cuvinte
Subiecte conexe
A Momentum-Based Variance-Reduced Algorithm for Federated Multiobjective Optimization
arXiv · 25 aug. 2026
Learning What to Evaluate: Correlation-Aware Decoupling for Multiobjective Bayesian Optimization
arXiv · 30 sept. 2026
Expected Hypervolume Maximization for Multiobjective Optimization under Uncertainties
arXiv · 18 sept. 2026
Solving Few-Shot Multiobjective Multitask Optimization via Iterative Sequential Transfer
arXiv · 12 sept. 2026