From 95e025b2b6e44e2eb7460b5a1fa088cb15165573 Mon Sep 17 00:00:00 2001 From: Melanie H Buehler Date: Tue, 19 Mar 2024 14:56:24 -0700 Subject: [PATCH] Improved text --- examples/Detecting_insecure_code.ipynb | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/examples/Detecting_insecure_code.ipynb b/examples/Detecting_insecure_code.ipynb index f0b220e4..d6ebd3db 100644 --- a/examples/Detecting_insecure_code.ipynb +++ b/examples/Detecting_insecure_code.ipynb @@ -919,7 +919,7 @@ "metadata": {}, "source": [ "### Experiment 3: KNN Few-shot\n", - "[This paper](https://arxiv.org/abs/2311.16452) describes an interesting technique called KNN-based few-shot example selection that led to a notable boost in performance. For this next experiment, instead of sampling shots at random, we calculated a similarity score between each candidate example and the input code and constructed shots from the most similar candidates (but still kept the scenarios distinct). We are using the RougeL fmeasure metric, but other metrics could be used too." + "[This paper](https://arxiv.org/abs/2311.16452) describes an interesting technique called KNN-based few-shot example selection that led to a notable boost in performance. For this next experiment, instead of sampling shots at random, we calculate a similarity score between each candidate example and the input code and construct shots from the most similar candidates (but still keep the scenarios distinct). This uses the RougeL fmeasure metric, but other metrics could be used too." ] }, {