False-science induction in autonomous scientific discovery
Abstract
Closed-loop discovery systems increasingly execute experiments and update decisions autonomously, turning record integrity into part of the experimental apparatus. We show that false-science induction arises when legitimate physical objects and measurements are paired incorrectly, driving neural surrogates to faithfully learn record-induced associations that do not correspond to the true object-outcome relationship while marginal data distributions remain unchanged. Across green fluorescent protein fitness and materials band-gap prediction loops, coherent paired misbinding systematically redirects experimental budgets toward low-performing basins, whereas same-volume random swaps have negligible effects. These observations identify error coherence, rather than raw error frequency, as the primary variable controlling this budget misallocation in the tested loops. The resulting binding identifiability boundary supports monitored-axis quarantines and feedback-conflict triage, which intercept over-concentrated proposals before execution and isolate the corrupted hypothesis axis.
Related
- From Automated Simulation to Autonomous Discovery: A Hierarchical Framework for Agentic Computational Materials Science
- Agentic programs: an emerging form of scientific software in computational materials science
- Mined from Scientific Literature: Process Schemas for Atomic Layer Deposition and Etching in Materials Science
- ADAPT for Science: A Multi-Agent System for Self-Evolving Scientific Optimization
- Decoding complexity through machine learning is redefining scientific discovery
BibTeX
@misc{liang2026falsescience,
title = {{False-science induction in autonomous scientific discovery}},
author = {Hanbing Liang and Fujun Liu},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.27883},
eprint = {2609.27883},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.27883},
}
RIS
TY - UNPB
TI - False-science induction in autonomous scientific discovery
AU - Hanbing Liang
AU - Fujun Liu
PY - 2026
DA - 2026/09/01
PB - arXiv
DO - 10.48550/arxiv.2609.27883
UR - https://arxiv.org/abs/2609.27883
AB - Closed-loop discovery systems increasingly execute experiments and update decisions autonomously, turning record integrity into part of the experimental apparatus. We show that false-science induction arises when legitimate physical objects and measurements are paired incorrectly, driving neural surrogates to faithfully learn record-induced associations that do not correspond to the true object-outcome relationship while marginal data distributions remain unchanged. Across green fluorescent protein fitness and materials band-gap prediction loops, coherent paired misbinding systematically redirects experimental budgets toward low-performing basins, whereas same-volume random swaps have negligible effects. These observations identify error coherence, rather than raw error frequency, as the primary variable controlling this budget misallocation in the tested loops. The resulting binding identifiability boundary supports monitored-axis quarantines and feedback-conflict triage, which intercept over-concentrated proposals before execution and isolate the corrupted hypothesis axis.
ER -