Well-Log Reconstruction with Pre-trained Geological Knowledge: A Knowledge-Conditioned Large-Language-Model Generative Agent
Organizations: College of Computer Science and Technology, China University of Petroleum (East China), Qingdao, 266580, China
Abstract
Missing well-log intervals make subsurface reconstruction ill-posed. Existing numerical regression models lack geological prior knowledge to constrain ambiguous inverse mappings. We propose GeoLog-Agent, a knowledge-conditioned generative agent using pre-trained large language models (LLMs) as a source of geological priors. A universal Geo-Harness translates LLM stratigraphic reasoning into formation-anchored petrophysical trajectories, feature-wise linear modulation, and conditional adversarial constraints. On the Norwegian North Sea benchmark, GeoLog-Agent achieves a mean of 0.9307 and a peak of 0.9713. It exceeds the nine-baseline average of 0.8387 by 0.0920, with a 30.97% average RMSE reduction, and exceeds the strongest baseline by 0.0274--0.0498 in , with 16.25%--24.05% lower RMSE. Ablations confirm the effectiveness and synergy of geological prior reasoning, multi-scale dilated generation, and conditional adversarial modules. Transferring the coupling mechanism to six conventional deep learning models improves their by 0.0162--0.0789 with cross-well stability. These results demonstrate how foundation-model domain reasoning can constrain deep generative reconstruction of missing petrophysical measurements.
Figures & tables
| Model | M50 Protocol (50% Contiguous Missing) | M100 Protocol (100% Full-Well Blind) | Overall Mean | ||||
|---|---|---|---|---|---|---|---|
| Mean | DEN RMSE | AC RMSE | Mean | DEN RMSE | AC RMSE | ||
| DLinear | 0.7893 | 0.1254 | 38.7793 | 0.5573 | 0.1754 | 59.7646 | 0.6733 |
| Autoformer | 0.8588 | 0.0773 | 41.7653 | 0.6347 | 0.1320 | 64.5900 | 0.7468 |
| LSTM | 0.8955 | 0.0750 | 34.5233 | 0.8054 | 0.1022 | 47.2646 | 0.8505 |
| BiLSTM | 0.9057 | 0.0715 | 32.8235 | 0.8220 | 0.0955 | 46.2017 | 0.8638 |
| TimesNet | 0.9166 | 0.0683 | 30.3651 | 0.8249 | 0.0970 | 44.9000 | 0.8708 |
| Scenario | Ablation Configuration | Well 1 (15-9-15) | Well 2 (15-9-17) | ||||
|---|---|---|---|---|---|---|---|
| DEN | AC | AC MAE | DEN | AC | AC MAE | ||
| M50 | 1. Unconditional GAN (w/o Geo-Harness Prior) | 0.9295 | 0.9171 | 15.2138 | 0.9403 | 0.9558 | 9.1682 |
| 2. GeoLog-Agent (Anonymized LLM Prior) | 0.9372 | 0.9161 | 15.5723 | 0.9316 | 0.9494 | 9.5689 | |
| 3. GeoLog-Agent (Numerical RF Prior) | 0.9385 | 0.9228 | 14.3854 | 0.9493 | 0.9652 | 8.0631 | |
| 4. GeoLog-Agent (w/o Multi-Scale Dilation, ) | 0.9268 | 0.9236 | 14.1076 | 0.9520 | 0.9616 | 8.1582 | |
| 5. GeoLog-Agent (w/o Adversarial Loss, Only) | 0.9412 | 0.9298 | 13.6587 | 0.9482 | 0.9590 | 8.2259 | |
| Protocol | Backbone Architecture | Vanilla | With Geo-Harness | Gain | RMSE Reduction (%) |
|---|---|---|---|---|---|
| M50 | LSTM | 0.8955 | 0.9424 | 0.0469 | 26.28% |
| BiLSTM | 0.9057 | 0.9453 | 0.0396 | 24.54% | |
| PatchTST | 0.9178 | 0.9479 | 0.0301 | 21.04% | |
| Informer | 0.9209 | 0.9490 | 0.0281 | 20.30% | |
| Transformer | 0.9192 | 0.9480 | 0.0288 | 20.43% | |
| GAN | 0.9357 | 0.9519 | 0.0162 | 14.17% |
Appendix figures & tables7 assets
Supplementary material from the paper’s appendix.
Appendix
| Protocol | Model | Bulk Density (DEN, ) | Compressional Transit Time (AC, ) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Well 1 (15-9-15) | Well 2 (15-9-17) | Well 1 (15-9-15) | Well 2 (15-9-17) | ||||||||||
| MAE | RMSE | MAE | RMSE | MAE | RMSE | MAE | RMSE | ||||||
| M50 | DLinear | 0.7343 | 0.0711 | 0.1215 | 0.6998 | 0.0743 | 0.1293 | 0.8712 | 20.9275 | 42.4008 | 0.8519 | 17.9351 | 35.1579 |
| Autoformer | 0.9057 | 0.0377 | 0.0724 | 0.8786 | 0.0446 | 0.0822 | 0.8936 | 17.8242 | 38.5345 | 0.7574 | 26.8737 | 44.9962 | |
| LSTM | 0.8974 | 0.0387 | 0.0755 | 0.9002 | 0.0361 | 0.0745 | 0.8846 | 18.7295 | 40.1211 | 0.8998 | 14.3537 | 28.9255 | |
| BiLSTM | 0.9099 | 0.0372 | 0.0708 | 0.9065 | 0.0353 | 0.0722 | 0.8872 | 18.8382 | 39.6770 | 0.9192 | 12.6309 | 25.9700 | |
| Model | Acoustic Impedance ( ) | Reservoir Porosity (p.u.) | Pay-Zone Delineation | |||
|---|---|---|---|---|---|---|
| RMSE | RMSE | Macro-F1 (%) | Accuracy (%) | |||
| DLinear | 0.5904 | 1.7655 | 0.3710 | 9.9259 | 38.39 | 71.99 |
| Autoformer | 0.6134 | 1.7342 | 0.7240 | 6.4740 | 57.38 | 78.68 |
| LSTM | 0.7654 | 1.3422 | 0.7998 | 5.6444 | 56.94 | 84.55 |
| BiLSTM | 0.7743 | 1.3072 | 0.8115 | 5.4497 | 58.19 | 85.20 |
| TimesNet | 0.7841 | 1.2753 | 0.8287 | 5.2316 | 59.44 | 85.67 |
Conditional Generative Modeling
Conditional generative modeling learns to produce data distributions steered by labels, inputs, text, or physical and structural constraints.
See all 419 papers