The electronic density of states (DOS) is conventionally computed from a relaxed crystal structure, which is unavailable for compounds that have been neither synthesized nor cataloged. Here we introduce DOSSIER ($\textbf{D}$ensity $\textbf{o}$f $\textbf{S}$tates from $\textbf{S}$to$\textbf{i}$chiometry with $\textbf{E}$ncoder $\textbf{R}$epresentations), a chemical language model that maps elemental composition directly to this spectrum. The encoder is pretrained by cross-modal knowledge distillation from a universal machine-learning interatomic potential; the transfer lowers the error by 11% when only 1,000 training examples are available. On the Mat2Spec benchmark, DOSSIER reaches a mean absolute error of 3.76 states eV$^{-1}$ against 3.64 for the best structure-aware model; on an extended Materials Project dataset, the predicted spectra yield band gaps and $\textit{d}$-band descriptors with useful accuracy. Screening 11,977 binary and 13,251 five-component high-entropy alloy compositions for a $\textit{d}$-projected DOS resembling that of NiPt$_{3}$ places known oxygen reduction electrocatalysts near the top of the ranking.
It is demonstrated that explicitly incorporating the domain knowledge of the interatomic bonds can significantly and systematically improve the prediction of $\beta$-ZnS/NaCl stability and an increasing amount of bond-informed recursion features improves the predictive accuracy.
Rohan D. Kumar, M. Forti, A. Naik et al.· 0 citations
The structural and optical properties of the doubly yttrium-doped boron cluster B$_{18}$Y$_2$ have been systematically investigated using density functional theory calculations. The lowest-energy structure was identified through extensive basin-hopping searches accelerated by a pre-trained MACE machine-learning potenti...
Machine-learning (ML) corrections to density functional approximations (DFAs) offer a route to improving electronic-structure predictions while retaining the efficiency of semilocal functionals. An important question is whether such corrections learn transferable improvements or instead compensate for errors specific t...
Abhishek Bhattacharjee, Kishan Kumar Mohanta, S. Jana et al.· 0 citations
The information bottleneck (IB) seeks a representation $U$ of a source $X$ that retains as much information as possible about a target $Y$, subject to a constraint on $I(U;X)$. A classical argument shows that it suffices to consider representations with at most $|\mathcal{X}|+1$ symbols, and this bound is known to be t...
Although $\mathbb{E}[\mathrm{dropout}(x)] = x$, here we show that $\mathbb{E}[\mathrm{LayerNorm}(\mathrm{dropout}(x))]$ is not equal to $\mathrm{LayerNorm}(x)$. Accordingly, the pattern of a Dropout layer followed by a LayerNorm, which is common to many AlphaFold2-based protein structure predictors, produces a systemat...
Isaac Ellmen, David Errington, Matthew I. J. Raybould et al.· 0 citations
We piece together textbook material to re-derive the parameter $\beta$ entering reduced density operators from quantum statistical mechanics. By re-deriving $\beta$, we show that the content of $\beta$ depends on the particular application of statistical mechanics. We focus on two different applications, namely, electr...
Ellen T. Ekstrøm, Jacob Pedersen, Ida-Marie Høyvik· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
