The efficiency–coverage trade-off of scientific priors in Bayesian search: a registered benchmark on metallic multi-element elasticity data
- 1Independent researcher
Abstract
Scientific priors speed up Bayesian optimisation in materials discovery, but a prior that steers a small budget toward the materials it favours may miss those it disfavours. We measured this in a benchmark with a registered analysis plan: 7,020 searches (3,000 of them over 3,550 metallic multielement Materials Project structures), for the shear modulus and the specific Young’s modulus, with a historical prior and a composition-only elemental-modulus heuristic, each mixed into expected improvement search with a fixed weight β. Coverage was the recall of the half of the top-5% materials that the prior itself scores lowest. Coverage loss at moderate weight was detected in 4 of 6 prior–objective settings. The elemental heuristic reduced the restricted mean number of acquisitions to a top-1% shear-modulus material by 13.9 at β = 0.25 with no detected loss of coverage, but its shear-modulus form, applied to the specific modulus, lost 44.0% or more of the baseline coverage at every nonzero weight. A weight both beneficial and safe under the registered criterion existed in 3 of 6 settings (2 under a post hoc joint error control), never above β = 0.25, and depended on the safety margin and on curation. At β = 0.5, in 5 of 6 settings the weighted search found more top-5% materials in total but, by its estimates, fewer of those its prior disfavours (a loss detected in 3): the prior redirected the search rather than reducing its yield. Faster discovery alone therefore does not show that a search keeps what its prior disfavours.
Related
- Accuracy, Reliability, and Efficiency of a MACE Machine-Learning Interatomic Potential for Aspirin: A Reproducible Benchmark for Molecular Simulation
- Resolving the Strength–Modulus–Elasticity Tradeoff in Elastomers Using Dual Phase‐Separated Nanodomains
- Flexoelectricity-driven softening of bend elasticity leads to spontaneous chiral symmetry breaking in a polar fluid
- Cellulose-Based Nanosponges with PEDOT:PSS as Pressure Sensors with Tunable Elasticity: A Study on Modulation of Mechanical Properties and Pressure Response
- Elasticity of polycrystalline davemaoite constrains its grain size in the lower mantle
BibTeX
@misc{asad2026efficiencycoverage,
title = {{The efficiency–coverage trade-off of scientific priors in Bayesian search: a registered benchmark on metallic multi-element elasticity data}},
author = {Muhammad Asad},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15011035/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15011035/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - The efficiency–coverage trade-off of scientific priors in Bayesian search: a registered benchmark on metallic multi-element elasticity data
AU - Muhammad Asad
PY - 2026
DA - 2026/10/26
PB - ChemRxiv
DO - 10.26434/chemrxiv.15011035/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15011035/v1
AB - Scientific priors speed up Bayesian optimisation in materials discovery, but a prior that steers a small budget toward the materials it favours may miss those it disfavours. We measured this in a benchmark with a registered analysis plan: 7,020 searches (3,000 of them over 3,550 metallic multielement Materials Project structures), for the shear modulus and the specific Young’s modulus, with a historical prior and a composition-only elemental-modulus heuristic, each mixed into expected improvement search with a fixed weight β. Coverage was the recall of the half of the top-5% materials that the prior itself scores lowest. Coverage loss at moderate weight was detected in 4 of 6 prior–objective settings. The elemental heuristic reduced the restricted mean number of acquisitions to a top-1% shear-modulus material by 13.9 at β = 0.25 with no detected loss of coverage, but its shear-modulus form, applied to the specific modulus, lost 44.0% or more of the baseline coverage at every nonzero weight. A weight both beneficial and safe under the registered criterion existed in 3 of 6 settings (2 under a post hoc joint error control), never above β = 0.25, and depended on the safety margin and on curation. At β = 0.5, in 5 of 6 settings the weighted search found more top-5% materials in total but, by its estimates, fewer of those its prior disfavours (a loss detected in 3): the prior redirected the search rather than reducing its yield. Faster discovery alone therefore does not show that a search keeps what its prior disfavours.
ER -