GENERATIVE-AI-ENGINEER-ASSOCIATE · Question #55
A Generative AI Engineer is building a system that will answer questions on currently unfolding news topics. As such, it pulls information from a variety of sources including articles and social…
The correct answer is B. Use only approved social media and news accounts to prevent unexpected toxic data from. The root cause of toxic outputs is toxic input data. By restricting data ingestion to only approved, vetted social media and news accounts, the system implements a proactive input guardrail that blocks toxic content before it ever reaches the LLM. This directly addresses the…
Question
A Generative AI Engineer is building a system that will answer questions on currently unfolding news topics. As such, it pulls information from a variety of sources including articles and social media posts. They are concerned about toxic posts on social media causing toxic outputs from their system. Which guardrail will limit toxic outputs?
Options
- AReduce the amount of context items the system will include in consideration for its response.
- BUse only approved social media and news accounts to prevent unexpected toxic data from
- CLog all LLM system responses and perform a batch toxicity analysis monthly.
- DImplement rate limiting.
How the community answered
(24 responses)- A4% (1)
- B83% (20)
- C8% (2)
- D4% (1)
Explanation
The root cause of toxic outputs is toxic input data. By restricting data ingestion to only approved, vetted social media and news accounts, the system implements a proactive input guardrail that blocks toxic content before it ever reaches the LLM. This directly addresses the pipeline at the source. Option A (reducing context items) doesn't filter for toxicity-it just uses less data. Option C (monthly batch analysis) is reactive logging, not a guardrail-harm has already occurred. Option D (rate limiting) controls request volume, not content quality, and has no effect on toxicity.
Topics
Community Discussion
No community discussion yet for this question.