Working effectively with LLMs means understanding what they are good at and where their limitations lie. This helps set realistic expectations and diagnose issues. Everything on this page applies whether the model is reading text, images, audio, video, or PDFs.
LLMs are complex statistical models that predict the most likely output based on patterns learned from training data. They don't "understand" in a human sense. When presented with your policy and content, they are finding the statistically most probable category based on the input.
Because language is often ambiguous, especially in moderation contexts, and their internal processes involve probability, LLMs can sometimes produce different outputs for the exact same input if the policy definition is not perfectly clear or the content is borderline. This variability, particularly with ambiguous policies, is a normal characteristic of LLMs, not necessarily a model error. It highlights areas where your policy needs greater precision.
You can "teach" LLMs by giving them better structure, boundaries, context, and information. In the app, that happens through your policy text, your examples, and the Instructions box at the top of your policy, where you can add context that applies to the whole policy.
Unlike humans, LLMs need structure to function reliably. How you structure and order policies is as important as the text of the policies themselves. A well-structured and logically formatted policy will perform much better than a raw human-readable policy. See ‣ for how to structure categories, examples, and exceptions.
LLMs rely heavily on the explicit instructions and definitions provided in your policy text. They cannot reliably infer intent, access real-world context outside the provided text, or apply external knowledge like local laws unless that information is clearly included or defined.
The app processes each piece of content independently. LLMs do not remember previous content they've classified for the same user or understand concepts like "repeated violations" unless that history or context is explicitly provided as part of the input for each piece of content.
There is also a limit on how much text a model can process in a single request — the "context window." Very long content (large documents, long threads) can exceed it and be truncated, which can mean missed violations. If your platform has very long content, consider splitting it or ask us about the right way to handle it.
LLMs can be inherently cautious or conservative when evaluating content, particularly in sensitive areas like adult content, violence, hate speech, or self-harm. While this caution is intended to prevent the model from generating harmful responses, it can result in over-flagging (false positives) when applied to moderation tasks, especially if your specific policy allows certain types of content within these sensitive domains.
LLMs are trained on vast internet data and can inherit societal biases present in that data. This can lead to biased outcomes in moderation — for example, disproportionately flagging content related to certain protected characteristics even when the content itself doesn't violate the policy.