AIF-C01 Fundamentals of Generative AI Practice Question

Exhibit

Refer to the exhibit.

```
{
  "modelId": "amazon.titan-text-lite-v1",
  "contentType": "application/json",
  "accept": "application/json",
  "body": {
    "inputText": "Summarize the following meeting notes: ...",
    "textGenerationConfig": {
      "maxTokenCount": 100,
      "stopSequences": [],
      "temperature": 0,
      "topP": 0.9
    }
  }
}
```

A developer is using the Amazon Bedrock InvokeModel API with the above request to summarize meeting notes. The response is a single word repeated many times. Which parameter is MOST likely causing this issue?

⚠ Common exam trap

AWS often tests the misconception that temperature only affects 'creativity' or 'randomness,' when in fact a temperature of 0 causes deterministic argmax selection, which can paradoxically produce repetitive or stuck outputs rather than simply 'less creative' text.

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

✓

temperature set to 0

A temperature of 0 forces the model to always select the highest-probability token at each step, which can lead to repetitive loops if the most likely token repeatedly points back to itself (e.g., the same word). This deterministic behavior eliminates randomness, causing the model to get stuck in a single-word cycle rather than generating diverse or coherent text.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✗
topP set to 0.9
Why it's wrong here
topP at 0.9 is a normal nucleus-sampling value that keeps most probability mass, so it does not by itself force repetition. It is tempting because sampling parameters govern token choice, which would be the correct area to tune when output is too random, but the repetition points to temperature or token limits.
✗
stopSequences is empty
Why it's wrong here
An empty stopSequences list simply means no custom stop strings are supplied; it does not cause looping output. It is tempting because stopping sequences do terminate generation, which would be the right control when you need output to halt at a delimiter, but repetition stems from sampling parameters instead.
✗
maxTokenCount set to 100
Why it's wrong here
maxTokenCount caps response length; a low value truncates output rather than repeating a word. It is tempting because token limits do shape responses, which would be the correct parameter when replies are cut off mid-sentence, but the looping here comes from sampling settings, not the cap.
✓
temperature set to 0
Why this is correct
Temperature at 0 makes the model greedy, always picking the highest-probability token, so a repetitive loop can dominate the output. It satisfies the stem's constraint by explaining the degenerate single-word repetition, though top-p or top-k sampling would restore diversity.

Go deeper

AIF-C01 Fundamentals of Generative AI Practice Question

Exhibit

Why each option matters

Option-by-option breakdown

Related to this question