Research assistant agent in Agentic AI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

How does a research assistant agent prioritize tasks?

A research assistant agent receives multiple tasks such as summarizing papers, extracting data, and generating reports. How does it decide which task to do first?

A It always does the easiest task first to finish quickly.

B It randomly picks any task without considering importance or deadlines.

C It waits for a human to assign the next task each time.

D It prioritizes tasks based on deadlines and relevance to the main research goal.

💡 Hint

Think about how a helpful assistant would manage multiple jobs efficiently.

A research assistant agent prioritizes tasks by considering deadlines and how important each task is to the main research goal. This helps it work efficiently and deliver useful results on time.

❓ Predict Output

intermediate

Output of a simple research assistant agent task queue

What will be the output of the following Python code simulating a research assistant agent's task queue?

Agentic AI

tasks = ['summarize paper', 'extract data', 'generate report']
completed = []

while tasks:
    task = tasks.pop(0)
    completed.append(f"Done: {task}")

print(completed)

A ['Done: summarize paper', 'Done: extract data', 'Done: generate report']

B ['Done: generate report', 'Done: extract data', 'Done: summarize paper']

C ['summarize paper', 'extract data', 'generate report']

D []

💡 Hint

Look at how tasks are removed and added to completed.

The code removes tasks from the front of the list and appends a 'Done:' message to completed in the same order, so the output list matches the original task order with 'Done:' prefix.

❓ Model Choice

advanced

Best model type for a research assistant agent summarizing papers

Which model type is most suitable for a research assistant agent tasked with summarizing scientific papers?

A A convolutional neural network (CNN) designed for image recognition.

B A recurrent neural network (RNN) or transformer model trained on text summarization.

C A clustering algorithm like K-means for grouping data points.

D A reinforcement learning model for playing games.

💡 Hint

Think about which model handles text and language well.

RNNs and transformer models are designed to process and generate text, making them ideal for summarizing papers. CNNs and clustering are not suited for text summarization tasks.

❓ Hyperparameter

advanced

Choosing hyperparameters for a research assistant agent's text summarization model

Which hyperparameter adjustment is most likely to improve the quality of summaries generated by a transformer-based research assistant agent?

A Increasing the learning rate drastically to speed up training.

B Reducing the number of attention heads to simplify the model.

C Increasing the maximum sequence length to capture more context.

D Decreasing the batch size to zero.

💡 Hint

Think about how the model understands longer texts.

Increasing the maximum sequence length allows the model to consider more of the input text at once, improving summary quality. Too high learning rate or zero batch size cause training issues, and fewer attention heads reduce model capacity.

❓ Metrics

expert

Evaluating a research assistant agent's summarization quality

A research assistant agent generates summaries of research papers. Which metric best measures how well the summaries capture the original content?

A BLEU score comparing generated summary to reference summary.

B Mean squared error between summary and original text embeddings.

C Accuracy of classification labels.

D Confusion matrix of predicted vs actual categories.

💡 Hint

Think about metrics used in language generation tasks.

BLEU score measures how closely the generated summary matches a reference summary, making it suitable for evaluating summarization quality. Accuracy and confusion matrix apply to classification, and mean squared error on embeddings is less interpretable here.