Internal workings of feed-forward neural networks
Bibliographic record
Abstract
This paper describes the authors' recent research on the theories about internal workings of feed-forward artificial neural networks (ANNs). Knowledge of the inner workings of ANNs progresses with the evolution of ANN modelling techniques. In the early stages, neural networks consisted of perceptrons and were easy to interpret. Since the emergence of backpropagation learning in the 1980s, ANN models have become much more complex. Even today, many users still consider ANN models as black-box type models. In this paper, a systematic approach is demonstrated to study the behaviours of ANN models in an attempt to open up the ANN "black box". The authors emphasise that feed-forward ANN models are functions, and use a graphical interpretation technique to open the ANN models and show the detailed activities of each inner component through several case studies. These studies include how the training process changes the internal components of an ANN model, how noise impacts the training process, and how the sensitivity of ANN models is affected by the training data. Key words: artificial neural network, ANN black box, artificial neuron, connection weight, activation function, knowledge extraction.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.017 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.003 | 0.006 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".