NCA-GENL Core Machine Learning and AI Knowledge Practice Question
A researcher is training a large language model and notices the training loss plateaus early while validation loss increases. What is the most likely cause, and which action should be taken?
⚠ Common exam trap
Candidates often misdiagnose increasing validation loss alongside a plateauing training loss as underfitting or a need for a higher learning rate, instead of recognizing classic overfitting.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
The model is overfitting; apply dropout or weight decay.
The symptoms described clearly indicate overfitting, where the model captures noise in the training set rather than generalizing to unseen data. In the context of large language models, this is a critical challenge. Implementing regularization techniques such as weight decay or dropout helps constrain model complexity, forcing it to learn more robust features rather than memorizing specific patterns, thereby improving overall model performance and generalizability.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
The model is underfitting; increase the number of hidden layers.
Why it's wrong here
Underfitting typically manifests as high bias, where both training and validation losses are high and stagnant. Increasing depth would exacerbate overfitting if the model is already memorizing the training data. The primary issue here is poor generalization, not insufficient model capacity to capture the underlying data patterns.
- ✗
The learning rate is too low; increase it to accelerate convergence.
Why it's wrong here
A very low learning rate would lead to slow convergence, making the loss decrease gradually rather than plateauing abruptly while validation loss rises. While adjusting the learning rate is standard, it does not directly address the divergence between training and validation error caused by model overfitting.
- ✓
The model is overfitting; apply dropout or weight decay.
Why this is correct
Overfitting occurs when the model complexity exceeds the information content in the training set. Dropout randomly disables neurons during training, preventing co-adaptation, while weight decay penalizes large weights. These techniques effectively reduce the variance of the model, forcing it to focus on generalized representations instead of training noise.
- ✗
The dataset is too small; reduce the batch size.
Why it's wrong here
Reducing batch size might introduce more stochasticity into the gradient estimates, but it does not inherently solve overfitting. While more data is ideal, the immediate fix for overfitting on an existing dataset is regularization. Changing batch size primarily affects memory utilization and the stability of the training process.
About these practice questions
Courseiva writes every NCA-GENL question from scratch — 367 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official NVIDIA exam blueprint
This NCA-GENL practice question is part of Courseiva's free NVIDIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the NCA-GENL exam.