---
res:
  bibo_abstract:
  - "Property prediction for molecules and materials is bottlenecked by label scarcity,
    and in this regime the operative failure mode is not a low mean but an unpredictable
    tail: individual training runs collapse. The three dominant molecular representations
    fail in complementary ways?physicochemical descriptors are exact but fixed, graph
    neural networks learn topology but degenerate under sparse supervision, and language
    models supply semantic context but cannot compute exact quantities? yet combining
    them is itself the learning problem, because modality-level scalar gating commits
    the whole model to one trust weight per source. We introduce TAME, a tri-modal
    encoder whose element-wise Mixture-of-Experts router assigns an independent modality
    mixture to every hidden coordinate, stabilized by a balance?entropy regularizer,
    a closed-form gate initialization (? = log Ï\x84 ) that is intended to remove
    router burn-in, and two-stage self-supervised graph pretraining. On scaffold-split
    BACE (â\x89\x88 1.5k molecules, 100 seeds) we evaluate two fusion topologies,
    each against its own graph-only control. In both, adding the text and descriptor
    experts contracts the seed-to-seed distribution of the threshold-free metrics?ROC-AUC
    Ï\x83 falls five-fold in the flat topology and 1.6-fold in the hierarchical one,
    whose control was already the tighter of the two?and the collapse tail reaching
    ROC-AUC â\x89\x88 0.45 that the weaker single-modality configurations carry is
    absent from the fused models while accuracy is retained. Each comparison is matched
    on everything but the fusion stage, isolating its effect. The router allocates
    a distinct mixture to each coordinate, which no scalar gate can represent, and
    the entropy weight moves that allocation continuously between graded mixing and
    a near-binary regime in which each coordinate is claimed outright by one modality.
    Fused-representation alignment tracks encoder quality without supervision, shifting
    from text to topology once pretraining lifts the graph embedding out of rank-1
    collapse. The same three experts map onto crystal graphs, composition?structure
    descriptors and literature-mined text. These three modalities recur unchanged
    across molecular and inorganic chemistry, so TAME is a single interpretable recipe
    for turning volatile low-data predictors into reliable, self-explaining ones,
    wherever data are scarce and every failed prediction costs a real experiment.@eng"
  bibo_authorlist:
  - foaf_Person:
      foaf_givenName: Robert P.
      foaf_name: Schiller, Robert P.
      foaf_surname: Schiller
  - foaf_Person:
      foaf_givenName: Christoph
      foaf_name: Weisser, Christoph
      foaf_surname: Weisser
      foaf_workInfoHomepage: http://www.librecat.org/personId=264138
    orcid: 0000-0003-0616-1027
    orcid_put_code_url: https://api.orcid.org/v2.0/0000-0003-0616-1027/work/226396719
  - foaf_Person:
      foaf_givenName: Klaus-Robert
      foaf_name: Müller, Klaus-Robert
      foaf_surname: Müller
  - foaf_Person:
      foaf_givenName: Christian
      foaf_name: Ochsenfeld, Christian
      foaf_surname: Ochsenfeld
  - foaf_Person:
      foaf_givenName: Parastoo
      foaf_name: Semnani, Parastoo
      foaf_surname: Semnani
  bibo_doi: 10.26434/chemrxiv.15008270/v2
  dct_date: 2026^xs_gYear
  dct_language: eng
  dct_title: 'TAME: Element-wise Mixture-of-Experts Fusion for Reliable and Interpretable
    Molecular Property Prediction@'
...