Performance of small basis set Hartree–Fock methods for modeling non-covalent interactions
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Abstract Non-covalent interactions (NCIs) play an essential role in (bio)chemistry. Wavefunction-based methods combined with large basis sets are able to accurately describe inter-and intra-molecular NCIs but are not practical for large molecular systems. Semi-empirical corrections have been developed recently that, when combined with Hartree–Fock (HF) and a small basis set, show promise in the ability to predict non-covalent binding and conformational energies over a wide range of systems. Compared to large-basis-set correlated wavefunction methods, small-basis-set HF methods significantly lower computational cost and are useful for modeling large molecular systems with sizes between many hundreds and a few thousand atoms. Using a large collection of non-covalent binding energies, conformational energies, and molecular deformation energies containing 105 880 entries, we provide a comprehensive evaluation of the performance of the minimal basis set (MINIX) HF method with three correction schemes: D3, 3c, and atom-centered potentials (ACPs). We also evaluate the performance of HF/6-31G* in combination with the D3 and ACP schemes. By comparing the three corrections, we analyze the strengths and weaknesses associated with each strategy in predicting NCIs. Our results show that D3 corrections alone do not offer significant improvements in the performance of HF/MINIX or HF/6-31G* and, in some cases, overestimate binding energies resulting in large errors when compared to the reference data. The correction strategies that offer the best reduction in the underlying errors of HF/MINIX and HF/6-31G* are shown to be 3c and ACP for HF/MINIX and ACP for HF/6-31G*.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it