diff --git a/website/docs/api/large-language-models.mdx b/website/docs/api/large-language-models.mdx
index f8404cb2e..55d137e21 100644
--- a/website/docs/api/large-language-models.mdx
+++ b/website/docs/api/large-language-models.mdx
@@ -752,6 +752,25 @@ supports `.yml`, `.yaml`, `.json` and `.jsonl`.
 path = "textcat_examples.json"
 ```
 
+If you want to perform few-shot learning with a binary classifier (i. e. a text
+either should or should not be assigned to a given class), you can provide
+positive and negative examples with answers of "POS" or "NEG". "POS" means that
+this example should be assigned the class label defined in the configuration,
+"NEG" means it shouldn't. E. g. for spam classification:
+
+```json
+[
+  {
+    "text": "You won the lottery! Wire a fee of 200$ to be able to withdraw your winnings.",
+    "answer": "POS"
+  },
+  {
+    "text": "Your order #123456789 has arrived",
+    "answer": "NEG"
+  }
+]
+```
+
 ### REL {id="rel"}
 
 The REL task extracts relations between named entities.