readme modified

Author: sindchad

implementations/implementation_d/01_intoduction_and_dataset_construction.ipynb

@@ -35,7 +35,7 @@
     "- Robustness to bias  \n",
     "- Alignment transparency \n",
     "\n",
-    "![Figure 1: Scalar Reward Model vs Generative Judge](assets/Figure1.png)\n",
+    "![Figure 1: Scalar Reward Model vs Generative Judge](./assets/Figure1.png)\n",
     "\n",
     "\n",
     "## Con-J Training Pipeline\n",
@@ -64,7 +64,7 @@
     "\n",
     "This directly optimizes the model to prefer correct judgments while maintaining generation quality.\n",
     "\n",
-    "![Figure 2: Con-J Pipeline](assets/Figure2.png)\n",
+    "![Figure 2: Con-J Pipeline](./assets/Figure2.png)\n",
     "\n",
     "## What This Notebook Does\n",
     "\n",