llm attr

Author: aobo-y

tutorials/LLama2-Captum-Demo.ipynb renamed to tutorials/Llama2_LLM_Attribution.ipynb

@@ -5,7 +5,7 @@
    "id": "9bdd6ea2",
    "metadata": {},
    "source": [
-    "# Use Captum LLM Attribution to understand Llama2\n",
+    "# Understanding Llama2 with Captum LLM Attribution\n",
     "\n",
     "In this tutorial, we will demonstrate the LLM attribution functionality introduced in Captum v0.7, which makes it a breeze to applying the attribution algorithms to interpret the large langague models (LLM) in text generation. The new functionalities include a series utilities that help you with many common tedious scaffolding required by LLMs like defining intepretable features in text input and handling the sequential predictions. You can also check our paper for more details https://arxiv.org/abs/2312.05491\n",
     "\n",
@@ -490,7 +490,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "faaebc70",
+   "id": "aa2739bf",
    "metadata": {},
    "source": [
     "Interestingly, we can see all these few-shot examples we choose actually make the model less likely to correctly label the given review as \"Positive\"."

website/pages/tutorials/index.js

@@ -85,6 +85,11 @@ class TutorialHome extends React.Component {
                 Using Captum and Integrated Gradients we interpret the output of several test questions and analyze the attribution scores
                 of the text and visual parts of the model. Find the tutorial <a href="Multimodal_VQA_Interpret">here</a>.
 
+                <h4>Understanding Llama2 with Captum LLM Attribution:</h4>
+                This tutorial demonstrates how to easily use the LLM attribution functionality to interpret the large langague models (LLM) in text generation.
+                It takes Llama2 as the example and shows the step-by-step improvements from the basic attribution setting to more advanced techniques.
+                Find the tutorial <a href="Llama2_LLM_Attribution">here</a>.
+
                 <h4>Interpreting question answering with BERT Part 1:</h4>
                 This tutorial demonstrates how to use Captum to interpret a BERT model for question answering.
                 We use a pre-trained model from Hugging Face fine-tuned on the SQUAD dataset and show how to use hooks to

website/tutorials.json

@@ -46,6 +46,10 @@
       "id": "Image_and_Text_Classification_LIME",
       "title": "Interpreting vision and text models with LIME"
     },
+    {
+      "id": "Llama2_LLM_Attribution",
+      "title": "Understanding Llama2 with Captum LLM Attribution"
+    },
     {
       "title": "Interpreting BERT",
       "children": [