Position: AI for Science Should Treat Measurement-to-Dataset Pipelines as Inference Components

Ling Zhan, Xiaoyao Yu, Tao Jia
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:172873-172895, 2026.

Abstract

AI for Science (AI4Science) workflows often treat the released dataset as a fixed interface to the underlying system. However, in domains relying on indirect observation, the learner observes a derivative representation produced by multi-stage measurement, reconstruction, and preprocessing pipelines. We argue that these measurement-to-dataset pipelines are inference components: treating their outputs as "given data" freezes an observation model and obscures uncertainty over feasible pipeline choices. We identify three failure modes arising from this "frozen lens": (C1) hidden hypothesis space, where the released dataset does not specify the pipeline configuration or its validity conditions; (C2) uncertified transportability, where a pipeline may be documented but its regime of validity is untested, so failures under distribution shift cannot be adjudicated; (C3) ungoverned multiplicity, where many defensible pipelines exist and dispersion is real but not propagated into uncertainty-aware evidence. We stress-test these claims with a large-scale neuroscience empirical audit, finding a survival rate of $\approx$ 0.0004% under a cross-dataset stability criterion. We call on the AI4Science community to make pipelines computable inference objects via domain-specific Computable Observation Frameworks. This shift enables quantifying pipeline adequacy and stability, converting implicit implementation choices into auditable, reproducible, and cumulative scientific evidence.

Cite this Paper


BibTeX
@InProceedings{pmlr-v306-zhan26g, title = {Position: {AI} for Science Should Treat Measurement-to-Dataset Pipelines as Inference Components}, author = {Zhan, Ling and Yu, Xiaoyao and Jia, Tao}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {172873--172895}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/zhan26g/zhan26g.pdf}, url = {https://proceedings.mlr.press/v306/zhan26g.html}, abstract = {AI for Science (AI4Science) workflows often treat the released dataset as a fixed interface to the underlying system. However, in domains relying on indirect observation, the learner observes a derivative representation produced by multi-stage measurement, reconstruction, and preprocessing pipelines. We argue that these measurement-to-dataset pipelines are inference components: treating their outputs as "given data" freezes an observation model and obscures uncertainty over feasible pipeline choices. We identify three failure modes arising from this "frozen lens": (C1) hidden hypothesis space, where the released dataset does not specify the pipeline configuration or its validity conditions; (C2) uncertified transportability, where a pipeline may be documented but its regime of validity is untested, so failures under distribution shift cannot be adjudicated; (C3) ungoverned multiplicity, where many defensible pipelines exist and dispersion is real but not propagated into uncertainty-aware evidence. We stress-test these claims with a large-scale neuroscience empirical audit, finding a survival rate of $\approx$ 0.0004% under a cross-dataset stability criterion. We call on the AI4Science community to make pipelines computable inference objects via domain-specific Computable Observation Frameworks. This shift enables quantifying pipeline adequacy and stability, converting implicit implementation choices into auditable, reproducible, and cumulative scientific evidence.} }
Endnote
%0 Conference Paper %T Position: AI for Science Should Treat Measurement-to-Dataset Pipelines as Inference Components %A Ling Zhan %A Xiaoyao Yu %A Tao Jia %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-zhan26g %I PMLR %P 172873--172895 %U https://proceedings.mlr.press/v306/zhan26g.html %V 306 %X AI for Science (AI4Science) workflows often treat the released dataset as a fixed interface to the underlying system. However, in domains relying on indirect observation, the learner observes a derivative representation produced by multi-stage measurement, reconstruction, and preprocessing pipelines. We argue that these measurement-to-dataset pipelines are inference components: treating their outputs as "given data" freezes an observation model and obscures uncertainty over feasible pipeline choices. We identify three failure modes arising from this "frozen lens": (C1) hidden hypothesis space, where the released dataset does not specify the pipeline configuration or its validity conditions; (C2) uncertified transportability, where a pipeline may be documented but its regime of validity is untested, so failures under distribution shift cannot be adjudicated; (C3) ungoverned multiplicity, where many defensible pipelines exist and dispersion is real but not propagated into uncertainty-aware evidence. We stress-test these claims with a large-scale neuroscience empirical audit, finding a survival rate of $\approx$ 0.0004% under a cross-dataset stability criterion. We call on the AI4Science community to make pipelines computable inference objects via domain-specific Computable Observation Frameworks. This shift enables quantifying pipeline adequacy and stability, converting implicit implementation choices into auditable, reproducible, and cumulative scientific evidence.
APA
Zhan, L., Yu, X. & Jia, T.. (2026). Position: AI for Science Should Treat Measurement-to-Dataset Pipelines as Inference Components. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:172873-172895 Available from https://proceedings.mlr.press/v306/zhan26g.html.

Related Material