diff --git a/model_evaluation/BiosBias Evaluation.ipynb b/model_evaluation/BiosBias Evaluation.ipynb
index 8fead0e2..fac15db4 100644
--- a/model_evaluation/BiosBias Evaluation.ipynb	
+++ b/model_evaluation/BiosBias Evaluation.ipynb	
@@ -9,7 +9,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -39,12 +39,12 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
-    "standard_data_path = 'gs://conversationai-models/biosbias/scored_data/standard_test.csv'\n",
-    "scrubbed_data_path = 'gs://conversationai-models/biosbias/scored_data/scrubbed_test.csv'\n",
+    "standard_data_path = 'gs://conversationai-models/biosbias/scored_data/test_standard_0409.csv'\n",
+    "scrubbed_data_path = 'gs://conversationai-models/biosbias/scored_data/test_very_scrubbed_0409.csv'\n",
     "\n",
     "perf_df = pd.read_csv(tf.gfile.Open(standard_data_path)).drop_duplicates(subset=['tokens'])\n",
     "scrubbed_df = pd.read_csv(tf.gfile.Open(scrubbed_data_path)).drop_duplicates(subset=['tokens'])"
@@ -52,18 +52,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "(19829, 201)\n",
-      "(19828, 69)\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "print(perf_df.shape)\n",
     "print(scrubbed_df.shape)"
@@ -71,7 +62,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -80,358 +71,18 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/html": [
-       "<div>\n",
-       "<style scoped>\n",
-       "    .dataframe tbody tr th:only-of-type {\n",
-       "        vertical-align: middle;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe tbody tr th {\n",
-       "        vertical-align: top;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe thead th {\n",
-       "        text-align: right;\n",
-       "    }\n",
-       "</style>\n",
-       "<table border=\"1\" class=\"dataframe\">\n",
-       "  <thead>\n",
-       "    <tr style=\"text-align: right;\">\n",
-       "      <th></th>\n",
-       "      <th>tokens</th>\n",
-       "      <th>gender</th>\n",
-       "      <th>label</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_0</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_1</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_2</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_3</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_4</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_5</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_6</th>\n",
-       "      <th>...</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_23</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_24</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_25</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_26</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_27</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_28</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_29</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_30</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_31</th>\n",
-       "      <th>tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_32</th>\n",
-       "    </tr>\n",
-       "  </thead>\n",
-       "  <tbody>\n",
-       "    <tr>\n",
-       "      <th>0</th>\n",
-       "      <td>[u'in', u'his', u'career', u'he', u'has', u're...</td>\n",
-       "      <td>M</td>\n",
-       "      <td>3</td>\n",
-       "      <td>0.000030</td>\n",
-       "      <td>1.481644e-16</td>\n",
-       "      <td>0.000003</td>\n",
-       "      <td>0.999415</td>\n",
-       "      <td>6.231773e-08</td>\n",
-       "      <td>1.657187e-08</td>\n",
-       "      <td>5.000493e-10</td>\n",
-       "      <td>...</td>\n",
-       "      <td>0.000237</td>\n",
-       "      <td>1.022816e-06</td>\n",
-       "      <td>0.000110</td>\n",
-       "      <td>0.001456</td>\n",
-       "      <td>9.247497e-09</td>\n",
-       "      <td>2.519182e-16</td>\n",
-       "      <td>0.000015</td>\n",
-       "      <td>3.130671e-04</td>\n",
-       "      <td>0.000080</td>\n",
-       "      <td>4.128487e-08</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>1</th>\n",
-       "      <td>[u'like', u'keats', u',', u'she', u'seeks', u'...</td>\n",
-       "      <td>F</td>\n",
-       "      <td>24</td>\n",
-       "      <td>0.000358</td>\n",
-       "      <td>5.320205e-12</td>\n",
-       "      <td>0.003868</td>\n",
-       "      <td>0.005619</td>\n",
-       "      <td>1.016141e-06</td>\n",
-       "      <td>1.004333e-04</td>\n",
-       "      <td>1.406151e-04</td>\n",
-       "      <td>...</td>\n",
-       "      <td>0.002092</td>\n",
-       "      <td>8.205463e-01</td>\n",
-       "      <td>0.005582</td>\n",
-       "      <td>0.008591</td>\n",
-       "      <td>4.002794e-04</td>\n",
-       "      <td>1.151956e-12</td>\n",
-       "      <td>0.002781</td>\n",
-       "      <td>5.614875e-04</td>\n",
-       "      <td>0.040721</td>\n",
-       "      <td>1.458319e-04</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>2</th>\n",
-       "      <td>[u'with', u'a', u'degree', u'in', u'both', u'f...</td>\n",
-       "      <td>F</td>\n",
-       "      <td>10</td>\n",
-       "      <td>0.000122</td>\n",
-       "      <td>2.086429e-14</td>\n",
-       "      <td>0.000262</td>\n",
-       "      <td>0.000889</td>\n",
-       "      <td>1.753830e-07</td>\n",
-       "      <td>4.022829e-03</td>\n",
-       "      <td>1.572652e-03</td>\n",
-       "      <td>...</td>\n",
-       "      <td>0.000025</td>\n",
-       "      <td>3.176962e-04</td>\n",
-       "      <td>0.002818</td>\n",
-       "      <td>0.000234</td>\n",
-       "      <td>1.174002e-05</td>\n",
-       "      <td>9.342651e-15</td>\n",
-       "      <td>0.000654</td>\n",
-       "      <td>1.156893e-05</td>\n",
-       "      <td>0.002540</td>\n",
-       "      <td>1.259932e-05</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>3</th>\n",
-       "      <td>[u'he', u'has', u'been', u'considered', u'as',...</td>\n",
-       "      <td>M</td>\n",
-       "      <td>18</td>\n",
-       "      <td>0.000187</td>\n",
-       "      <td>1.589816e-11</td>\n",
-       "      <td>0.014234</td>\n",
-       "      <td>0.003778</td>\n",
-       "      <td>9.938718e-06</td>\n",
-       "      <td>1.228533e-03</td>\n",
-       "      <td>2.325906e-02</td>\n",
-       "      <td>...</td>\n",
-       "      <td>0.002938</td>\n",
-       "      <td>5.874482e-03</td>\n",
-       "      <td>0.054547</td>\n",
-       "      <td>0.003046</td>\n",
-       "      <td>3.182003e-03</td>\n",
-       "      <td>1.945734e-11</td>\n",
-       "      <td>0.000654</td>\n",
-       "      <td>2.747838e-03</td>\n",
-       "      <td>0.035171</td>\n",
-       "      <td>9.126107e-05</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>4</th>\n",
-       "      <td>[u'his', u'research', u'interests', u'include'...</td>\n",
-       "      <td>M</td>\n",
-       "      <td>25</td>\n",
-       "      <td>0.000003</td>\n",
-       "      <td>9.075201e-14</td>\n",
-       "      <td>0.000627</td>\n",
-       "      <td>0.000022</td>\n",
-       "      <td>2.752427e-08</td>\n",
-       "      <td>2.030256e-08</td>\n",
-       "      <td>1.343099e-06</td>\n",
-       "      <td>...</td>\n",
-       "      <td>0.000022</td>\n",
-       "      <td>9.040638e-07</td>\n",
-       "      <td>0.996070</td>\n",
-       "      <td>0.000824</td>\n",
-       "      <td>2.524824e-11</td>\n",
-       "      <td>1.079238e-14</td>\n",
-       "      <td>0.000436</td>\n",
-       "      <td>9.989949e-07</td>\n",
-       "      <td>0.000881</td>\n",
-       "      <td>1.911942e-08</td>\n",
-       "    </tr>\n",
-       "  </tbody>\n",
-       "</table>\n",
-       "<p>5 rows × 270 columns</p>\n",
-       "</div>"
-      ],
-      "text/plain": [
-       "                                              tokens gender  label  \\\n",
-       "0  [u'in', u'his', u'career', u'he', u'has', u're...      M      3   \n",
-       "1  [u'like', u'keats', u',', u'she', u'seeks', u'...      F     24   \n",
-       "2  [u'with', u'a', u'degree', u'in', u'both', u'f...      F     10   \n",
-       "3  [u'he', u'has', u'been', u'considered', u'as',...      M     18   \n",
-       "4  [u'his', u'research', u'interests', u'include'...      M     25   \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_0  \\\n",
-       "0                                           0.000030                           \n",
-       "1                                           0.000358                           \n",
-       "2                                           0.000122                           \n",
-       "3                                           0.000187                           \n",
-       "4                                           0.000003                           \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_1  \\\n",
-       "0                                       1.481644e-16                           \n",
-       "1                                       5.320205e-12                           \n",
-       "2                                       2.086429e-14                           \n",
-       "3                                       1.589816e-11                           \n",
-       "4                                       9.075201e-14                           \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_2  \\\n",
-       "0                                           0.000003                           \n",
-       "1                                           0.003868                           \n",
-       "2                                           0.000262                           \n",
-       "3                                           0.014234                           \n",
-       "4                                           0.000627                           \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_3  \\\n",
-       "0                                           0.999415                           \n",
-       "1                                           0.005619                           \n",
-       "2                                           0.000889                           \n",
-       "3                                           0.003778                           \n",
-       "4                                           0.000022                           \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_4  \\\n",
-       "0                                       6.231773e-08                           \n",
-       "1                                       1.016141e-06                           \n",
-       "2                                       1.753830e-07                           \n",
-       "3                                       9.938718e-06                           \n",
-       "4                                       2.752427e-08                           \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_5  \\\n",
-       "0                                       1.657187e-08                           \n",
-       "1                                       1.004333e-04                           \n",
-       "2                                       4.022829e-03                           \n",
-       "3                                       1.228533e-03                           \n",
-       "4                                       2.030256e-08                           \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247_6  \\\n",
-       "0                                       5.000493e-10                           \n",
-       "1                                       1.406151e-04                           \n",
-       "2                                       1.572652e-03                           \n",
-       "3                                       2.325906e-02                           \n",
-       "4                                       1.343099e-06                           \n",
-       "\n",
-       "                                      ...                                      \\\n",
-       "0                                     ...                                       \n",
-       "1                                     ...                                       \n",
-       "2                                     ...                                       \n",
-       "3                                     ...                                       \n",
-       "4                                     ...                                       \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_23  \\\n",
-       "0                                           0.000237                            \n",
-       "1                                           0.002092                            \n",
-       "2                                           0.000025                            \n",
-       "3                                           0.002938                            \n",
-       "4                                           0.000022                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_24  \\\n",
-       "0                                       1.022816e-06                            \n",
-       "1                                       8.205463e-01                            \n",
-       "2                                       3.176962e-04                            \n",
-       "3                                       5.874482e-03                            \n",
-       "4                                       9.040638e-07                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_25  \\\n",
-       "0                                           0.000110                            \n",
-       "1                                           0.005582                            \n",
-       "2                                           0.002818                            \n",
-       "3                                           0.054547                            \n",
-       "4                                           0.996070                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_26  \\\n",
-       "0                                           0.001456                            \n",
-       "1                                           0.008591                            \n",
-       "2                                           0.000234                            \n",
-       "3                                           0.003046                            \n",
-       "4                                           0.000824                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_27  \\\n",
-       "0                                       9.247497e-09                            \n",
-       "1                                       4.002794e-04                            \n",
-       "2                                       1.174002e-05                            \n",
-       "3                                       3.182003e-03                            \n",
-       "4                                       2.524824e-11                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_28  \\\n",
-       "0                                       2.519182e-16                            \n",
-       "1                                       1.151956e-12                            \n",
-       "2                                       9.342651e-15                            \n",
-       "3                                       1.945734e-11                            \n",
-       "4                                       1.079238e-14                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_29  \\\n",
-       "0                                           0.000015                            \n",
-       "1                                           0.002781                            \n",
-       "2                                           0.000654                            \n",
-       "3                                           0.000654                            \n",
-       "4                                           0.000436                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_30  \\\n",
-       "0                                       3.130671e-04                            \n",
-       "1                                       5.614875e-04                            \n",
-       "2                                       1.156893e-05                            \n",
-       "3                                       2.747838e-03                            \n",
-       "4                                       9.989949e-07                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_31  \\\n",
-       "0                                           0.000080                            \n",
-       "1                                           0.040721                            \n",
-       "2                                           0.002540                            \n",
-       "3                                           0.035171                            \n",
-       "4                                           0.000881                            \n",
-       "\n",
-       "   tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954_32  \n",
-       "0                                       4.128487e-08                           \n",
-       "1                                       1.458319e-04                           \n",
-       "2                                       1.259932e-05                           \n",
-       "3                                       9.126107e-05                           \n",
-       "4                                       1.911942e-08                           \n",
-       "\n",
-       "[5 rows x 270 columns]"
-      ]
-     },
-     "execution_count": 5,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "df.head()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "(19829, 270)"
-      ]
-     },
-     "execution_count": 6,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
     "df.shape"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "(19828, 270)\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "df = df.dropna()\n",
     "print(df.shape)"
@@ -446,7 +97,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -458,7 +109,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -470,25 +121,27 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "MODEL_NAMES = {\n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113247': 'debiased_tuned',\n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113241': 'debiased_untuned',\n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113114': 'glove_tuned',\n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113106': 'glove_untuned',\n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_163707': 'strong_tuned',\n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_163723': 'strong_untuned',\n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_113045': 'scrubbed_tuned', \n",
-    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190315_112954': 'scrubbed_untuned',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190328_103117': 'glove',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_174837': 'debiased_tolga',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_174941': 'debiased_biosbias',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_175003': 'strongdebias_1',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_175019': 'strongdebias_2',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_175034': 'strongdebias_3',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_175055': 'strongdebias_4', \n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_175113': 'no_equalize',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_175131': 'no_proj',\n",
+    "    'tf_trainer_tf_gru_attention_multiclass_biosbias_glove:v_20190410_175254': 'very_scrubbed'\n",
     "}"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -497,7 +150,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -507,12 +160,12 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "# Labels with either gender having too few examples\n",
-    "bad_labels = df.groupby('label').gender.value_counts().reset_index(name = 'count').query('count < 5').label.values\n",
+    "bad_labels = df.groupby('label').gender.value_counts().reset_index(name = 'count').query('count < 4').label.values\n",
     "assert len(bad_labels) == 0"
    ]
   },
@@ -525,24 +178,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Accuracy for model debiased_tuned: 0.815715150293\n",
-      "Accuracy for model debiased_untuned: 0.814807343151\n",
-      "Accuracy for model strong_tuned: 0.819144643938\n",
-      "Accuracy for model glove_tuned: 0.82045592092\n",
-      "Accuracy for model strong_untuned: 0.813596933629\n",
-      "Accuracy for model glove_untuned: 0.816169053863\n",
-      "Accuracy for model scrubbed_tuned: 0.819144643938\n",
-      "Accuracy for model scrubbed_untuned: 0.800534597539\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "accuracy_list = []\n",
     "for _model in MODEL_NAMES:\n",
@@ -561,7 +199,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -571,7 +209,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -581,7 +219,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -593,7 +231,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -602,43 +240,77 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "## WITH THRESHOLD\n",
+    "\n",
+    "# def compute_tpr(df, _class, _model, threshold = 0.5):\n",
+    "#     tpr = metrics.recall_score(df['label_{}'.format(_class)],\n",
+    "#                                df['{}_{}'.format(_model,_class)] > threshold)\n",
+    "#     return tpr\n",
+    "    \n",
+    "# def compute_tpr_tnr(df, _class, _model, threshold = 0.5):\n",
+    "#     #cm = metrics.confusion_matrix(df['label_{}'.format(_class)],\n",
+    "#     #                              df['{}_{}'.format(_model,_class)] > threshold)\n",
+    "#     cm = pd.crosstab(df['label_{}'.format(_class)], df['{}_{}'.format(_model,_class)] > threshold)\n",
+    "#     #display(cm)\n",
+    "#     if cm.shape[0] > 1:\n",
+    "#         tn = cm.iloc[0,0]\n",
+    "#         fp = cm.iloc[0,1]\n",
+    "#         fn = cm.iloc[1,0]\n",
+    "#         tp = cm.iloc[1,1]\n",
+    "#         tpr = tp/(tp+fn)\n",
+    "#         tnr = tn/(tn+fp)\n",
+    "#     else:\n",
+    "#         tpr = 0\n",
+    "#         tnr = 1\n",
+    "#     return tpr, tnr"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
-    "def compute_tpr(df, _class, _model, threshold = 0.5):\n",
+    "def compute_tpr(df, _class, _model, threshold=None):    \n",
     "    tpr = metrics.recall_score(df['label_{}'.format(_class)],\n",
-    "                               df['{}_{}'.format(_model,_class)] > threshold)\n",
+    "                               df['{}_class'.format(_model)] == _class)\n",
     "    return tpr\n",
+    "\n",
+    "def compute_tpr_tnr(df, _class, _model, threshold=None):\n",
     "    \n",
-    "def compute_tpr_by_gender(df, _class, _model, threshold = 0.5):\n",
-    "    tpr_m = compute_tpr(df.query('gender == \"M\"'), _class, _model, threshold)\n",
-    "    tpr_f = compute_tpr(df.query('gender == \"F\"'), _class, _model, threshold)\n",
-    "    return {'M': tpr_m, 'F': tpr_f}"
+    "    true_col = 'label_{}'.format(_class)\n",
+    "    pred_col = '{}_class'.format(_model)\n",
+    "    tn = len(df.loc[(df[true_col] == False) & (df[pred_col] != _class)])\n",
+    "    fp = len(df.loc[(df[true_col] == False) & (df[pred_col] ==_class)])\n",
+    "    fn = len(df.loc[(df[true_col] == True) & (df[pred_col] != _class)])\n",
+    "    tp = len(df.loc[(df[true_col] == True) & (df[pred_col] ==_class)])\n",
+    "\n",
+    "    if tp + fn == 0:\n",
+    "        raise ValueError('class {} has no positive - impossible to define tpr'.format(_class))\n",
+    "\n",
+    "    if tn + fp == 0:\n",
+    "        tpr = 1\n",
+    "        tnr = 0\n",
+    "    else:\n",
+    "        tpr = tp/(tp+fn)    \n",
+    "        tnr = tn/(tn+fp)\n",
+    "    return tpr, tnr"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
-    "def compute_tpr_tnr(df, _class, _model, threshold = 0.5):\n",
-    "    #cm = metrics.confusion_matrix(df['label_{}'.format(_class)],\n",
-    "    #                              df['{}_{}'.format(_model,_class)] > threshold)\n",
-    "    cm = pd.crosstab(df['label_{}'.format(_class)], df['{}_{}'.format(_model,_class)] > threshold)\n",
-    "    #display(cm)\n",
-    "    if cm.shape[0] > 1:\n",
-    "        tn = cm.iloc[0,0]\n",
-    "        fp = cm.iloc[0,1]\n",
-    "        fn = cm.iloc[1,0]\n",
-    "        tp = cm.iloc[1,1]\n",
-    "        tpr = tp/(tp+fn)\n",
-    "        tnr = tn/(tn+fp)\n",
-    "    else:\n",
-    "        tpr = 0\n",
-    "        tnr = 1\n",
-    "    return tpr, tnr\n",
+    "def compute_tpr_by_gender(df, _class, _model, threshold = 0.5):\n",
+    "    tpr_m = compute_tpr(df.query('gender == \"M\"'), _class, _model, threshold)\n",
+    "    tpr_f = compute_tpr(df.query('gender == \"F\"'), _class, _model, threshold)\n",
+    "    return {'M': tpr_m, 'F': tpr_f}\n",
     "\n",
     "def compute_tr_by_gender(df, _class, _model, threshold = 0.5):\n",
     "    tpr_m, tnr_m = compute_tpr_tnr(df.query('gender == \"M\"'), _class, _model, threshold)\n",
@@ -648,11 +320,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "for _class in CLASS_NAMES:\n",
+    "    true_col = 'label_{}'.format(_class)\n",
+    "    if len(df.loc[(df[true_col] == True)]) == 0:\n",
+    "        continue\n",
     "    for _model in MODEL_NAMES:\n",
     "        tpr_1 = compute_tpr(df, _class, _model)\n",
     "        tpr_2, _ = compute_tpr_tnr(df, _class, _model)\n",
@@ -662,7 +337,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 22,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -686,7 +361,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 23,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -695,7 +370,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 24,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -709,7 +384,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 25,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -718,92 +393,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 26,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/html": [
-       "<div>\n",
-       "<style scoped>\n",
-       "    .dataframe tbody tr th:only-of-type {\n",
-       "        vertical-align: middle;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe tbody tr th {\n",
-       "        vertical-align: top;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe thead th {\n",
-       "        text-align: right;\n",
-       "    }\n",
-       "</style>\n",
-       "<table border=\"1\" class=\"dataframe\">\n",
-       "  <thead>\n",
-       "    <tr style=\"text-align: right;\">\n",
-       "      <th></th>\n",
-       "      <th>frac_female</th>\n",
-       "    </tr>\n",
-       "  </thead>\n",
-       "  <tbody>\n",
-       "    <tr>\n",
-       "      <th>frac_female</th>\n",
-       "      <td>1.000000</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>debiased_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.442007</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>debiased_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.738924</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>strong_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.645211</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>glove_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.711791</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>strong_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.627283</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>glove_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.647357</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>scrubbed_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.341492</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>scrubbed_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.520765</td>\n",
-       "    </tr>\n",
-       "  </tbody>\n",
-       "</table>\n",
-       "</div>"
-      ],
-      "text/plain": [
-       "                                 frac_female\n",
-       "frac_female                         1.000000\n",
-       "debiased_tuned_tpr_gender_gap       0.442007\n",
-       "debiased_untuned_tpr_gender_gap     0.738924\n",
-       "strong_tuned_tpr_gender_gap         0.645211\n",
-       "glove_tuned_tpr_gender_gap          0.711791\n",
-       "strong_untuned_tpr_gender_gap       0.627283\n",
-       "glove_untuned_tpr_gender_gap        0.647357\n",
-       "scrubbed_tuned_tpr_gender_gap       0.341492\n",
-       "scrubbed_untuned_tpr_gender_gap     0.520765"
-      ]
-     },
-     "execution_count": 26,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
     "results_df[['frac_female']+['{}_tpr_gender_gap'.format(_model) for _model in MODEL_NAMES.values()]].corr()[['frac_female']]\n",
     "    "
@@ -811,7 +403,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 27,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -821,7 +413,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 28,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -831,927 +423,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 29,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/html": [
-       "<div>\n",
-       "<style scoped>\n",
-       "    .dataframe tbody tr th:only-of-type {\n",
-       "        vertical-align: middle;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe tbody tr th {\n",
-       "        vertical-align: top;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe thead th {\n",
-       "        text-align: right;\n",
-       "    }\n",
-       "</style>\n",
-       "<table border=\"1\" class=\"dataframe\">\n",
-       "  <thead>\n",
-       "    <tr style=\"text-align: right;\">\n",
-       "      <th></th>\n",
-       "      <th>label_profession</th>\n",
-       "      <th>frac_female</th>\n",
-       "      <th>debiased_tuned_tpr_gender_gap</th>\n",
-       "      <th>debiased_untuned_tpr_gender_gap</th>\n",
-       "      <th>strong_tuned_tpr_gender_gap</th>\n",
-       "      <th>glove_tuned_tpr_gender_gap</th>\n",
-       "      <th>strong_untuned_tpr_gender_gap</th>\n",
-       "      <th>glove_untuned_tpr_gender_gap</th>\n",
-       "      <th>scrubbed_tuned_tpr_gender_gap</th>\n",
-       "      <th>scrubbed_untuned_tpr_gender_gap</th>\n",
-       "      <th>debiased_tuned_tnr_gender_gap</th>\n",
-       "      <th>debiased_untuned_tnr_gender_gap</th>\n",
-       "      <th>strong_tuned_tnr_gender_gap</th>\n",
-       "      <th>glove_tuned_tnr_gender_gap</th>\n",
-       "      <th>strong_untuned_tnr_gender_gap</th>\n",
-       "      <th>glove_untuned_tnr_gender_gap</th>\n",
-       "      <th>scrubbed_tuned_tnr_gender_gap</th>\n",
-       "      <th>scrubbed_untuned_tnr_gender_gap</th>\n",
-       "    </tr>\n",
-       "  </thead>\n",
-       "  <tbody>\n",
-       "    <tr>\n",
-       "      <th>13</th>\n",
-       "      <td>nurse</td>\n",
-       "      <td>0.913333</td>\n",
-       "      <td>0.103781</td>\n",
-       "      <td>0.094142</td>\n",
-       "      <td>0.103406</td>\n",
-       "      <td>0.065413</td>\n",
-       "      <td>0.026483</td>\n",
-       "      <td>0.074677</td>\n",
-       "      <td>0.033034</td>\n",
-       "      <td>0.056990</td>\n",
-       "      <td>-0.013451</td>\n",
-       "      <td>-0.009189</td>\n",
-       "      <td>-0.006368</td>\n",
-       "      <td>-0.009822</td>\n",
-       "      <td>-0.004472</td>\n",
-       "      <td>-0.008299</td>\n",
-       "      <td>-0.008484</td>\n",
-       "      <td>-0.007496</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>7</th>\n",
-       "      <td>dietitian</td>\n",
-       "      <td>0.900000</td>\n",
-       "      <td>0.015873</td>\n",
-       "      <td>0.259259</td>\n",
-       "      <td>0.052910</td>\n",
-       "      <td>0.116402</td>\n",
-       "      <td>0.164021</td>\n",
-       "      <td>0.174603</td>\n",
-       "      <td>0.079365</td>\n",
-       "      <td>0.074074</td>\n",
-       "      <td>-0.002258</td>\n",
-       "      <td>-0.001876</td>\n",
-       "      <td>-0.001783</td>\n",
-       "      <td>-0.001969</td>\n",
-       "      <td>-0.001855</td>\n",
-       "      <td>-0.002917</td>\n",
-       "      <td>-0.001989</td>\n",
-       "      <td>-0.001423</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>27</th>\n",
-       "      <td>yoga_teacher</td>\n",
-       "      <td>0.896552</td>\n",
-       "      <td>0.017094</td>\n",
-       "      <td>-0.021368</td>\n",
-       "      <td>-0.008547</td>\n",
-       "      <td>0.188034</td>\n",
-       "      <td>0.012821</td>\n",
-       "      <td>-0.047009</td>\n",
-       "      <td>0.089744</td>\n",
-       "      <td>-0.021368</td>\n",
-       "      <td>-0.001124</td>\n",
-       "      <td>-0.001105</td>\n",
-       "      <td>-0.001012</td>\n",
-       "      <td>-0.001012</td>\n",
-       "      <td>-0.000900</td>\n",
-       "      <td>-0.001217</td>\n",
-       "      <td>-0.001441</td>\n",
-       "      <td>-0.000808</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>15</th>\n",
-       "      <td>paralegal</td>\n",
-       "      <td>0.857143</td>\n",
-       "      <td>0.393939</td>\n",
-       "      <td>0.454545</td>\n",
-       "      <td>0.242424</td>\n",
-       "      <td>0.227273</td>\n",
-       "      <td>0.287879</td>\n",
-       "      <td>0.409091</td>\n",
-       "      <td>0.318182</td>\n",
-       "      <td>0.242424</td>\n",
-       "      <td>-0.000208</td>\n",
-       "      <td>-0.001735</td>\n",
-       "      <td>-0.000004</td>\n",
-       "      <td>0.000185</td>\n",
-       "      <td>0.000054</td>\n",
-       "      <td>-0.000748</td>\n",
-       "      <td>0.000073</td>\n",
-       "      <td>-0.000170</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>12</th>\n",
-       "      <td>model</td>\n",
-       "      <td>0.819407</td>\n",
-       "      <td>0.555725</td>\n",
-       "      <td>0.568883</td>\n",
-       "      <td>0.563826</td>\n",
-       "      <td>0.558769</td>\n",
-       "      <td>0.529409</td>\n",
-       "      <td>0.585330</td>\n",
-       "      <td>0.492488</td>\n",
-       "      <td>0.532944</td>\n",
-       "      <td>-0.000361</td>\n",
-       "      <td>-0.001767</td>\n",
-       "      <td>-0.000536</td>\n",
-       "      <td>-0.001274</td>\n",
-       "      <td>0.001565</td>\n",
-       "      <td>-0.000913</td>\n",
-       "      <td>-0.000300</td>\n",
-       "      <td>0.000077</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>10</th>\n",
-       "      <td>interior_designer</td>\n",
-       "      <td>0.742424</td>\n",
-       "      <td>0.096038</td>\n",
-       "      <td>0.295318</td>\n",
-       "      <td>0.180072</td>\n",
-       "      <td>0.060024</td>\n",
-       "      <td>0.200480</td>\n",
-       "      <td>0.141657</td>\n",
-       "      <td>0.080432</td>\n",
-       "      <td>0.100840</td>\n",
-       "      <td>0.000166</td>\n",
-       "      <td>-0.000577</td>\n",
-       "      <td>-0.000207</td>\n",
-       "      <td>0.000055</td>\n",
-       "      <td>-0.000002</td>\n",
-       "      <td>-0.000988</td>\n",
-       "      <td>0.000313</td>\n",
-       "      <td>0.000090</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>22</th>\n",
-       "      <td>psychologist</td>\n",
-       "      <td>0.615632</td>\n",
-       "      <td>-0.003880</td>\n",
-       "      <td>0.014645</td>\n",
-       "      <td>0.001366</td>\n",
-       "      <td>-0.067831</td>\n",
-       "      <td>-0.011883</td>\n",
-       "      <td>-0.019193</td>\n",
-       "      <td>-0.024338</td>\n",
-       "      <td>0.019479</td>\n",
-       "      <td>-0.005905</td>\n",
-       "      <td>-0.004092</td>\n",
-       "      <td>-0.004249</td>\n",
-       "      <td>-0.001270</td>\n",
-       "      <td>-0.006115</td>\n",
-       "      <td>-0.005507</td>\n",
-       "      <td>-0.002363</td>\n",
-       "      <td>-0.004891</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>26</th>\n",
-       "      <td>teacher</td>\n",
-       "      <td>0.591990</td>\n",
-       "      <td>0.144606</td>\n",
-       "      <td>0.177934</td>\n",
-       "      <td>0.196176</td>\n",
-       "      <td>0.200904</td>\n",
-       "      <td>0.122057</td>\n",
-       "      <td>0.162972</td>\n",
-       "      <td>0.164425</td>\n",
-       "      <td>0.122142</td>\n",
-       "      <td>-0.002170</td>\n",
-       "      <td>-0.002134</td>\n",
-       "      <td>-0.002453</td>\n",
-       "      <td>-0.004788</td>\n",
-       "      <td>-0.000893</td>\n",
-       "      <td>-0.002372</td>\n",
-       "      <td>-0.003146</td>\n",
-       "      <td>-0.000795</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>17</th>\n",
-       "      <td>personal_trainer</td>\n",
-       "      <td>0.583333</td>\n",
-       "      <td>-0.028571</td>\n",
-       "      <td>-0.007143</td>\n",
-       "      <td>0.042857</td>\n",
-       "      <td>0.078571</td>\n",
-       "      <td>0.014286</td>\n",
-       "      <td>-0.121429</td>\n",
-       "      <td>0.064286</td>\n",
-       "      <td>-0.021429</td>\n",
-       "      <td>-0.000298</td>\n",
-       "      <td>0.000296</td>\n",
-       "      <td>-0.000205</td>\n",
-       "      <td>-0.000242</td>\n",
-       "      <td>-0.000520</td>\n",
-       "      <td>-0.000354</td>\n",
-       "      <td>-0.000854</td>\n",
-       "      <td>-0.000316</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>11</th>\n",
-       "      <td>journalist</td>\n",
-       "      <td>0.489489</td>\n",
-       "      <td>-0.008312</td>\n",
-       "      <td>0.053362</td>\n",
-       "      <td>0.009611</td>\n",
-       "      <td>0.062468</td>\n",
-       "      <td>0.033670</td>\n",
-       "      <td>0.074991</td>\n",
-       "      <td>-0.001792</td>\n",
-       "      <td>0.036690</td>\n",
-       "      <td>0.004109</td>\n",
-       "      <td>-0.000966</td>\n",
-       "      <td>0.001937</td>\n",
-       "      <td>-0.000169</td>\n",
-       "      <td>0.001959</td>\n",
-       "      <td>-0.000280</td>\n",
-       "      <td>0.003475</td>\n",
-       "      <td>0.000804</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>19</th>\n",
-       "      <td>physician</td>\n",
-       "      <td>0.485414</td>\n",
-       "      <td>0.019000</td>\n",
-       "      <td>0.018553</td>\n",
-       "      <td>0.032291</td>\n",
-       "      <td>0.041488</td>\n",
-       "      <td>0.037069</td>\n",
-       "      <td>0.002383</td>\n",
-       "      <td>0.063199</td>\n",
-       "      <td>0.021449</td>\n",
-       "      <td>0.009150</td>\n",
-       "      <td>0.007732</td>\n",
-       "      <td>0.007114</td>\n",
-       "      <td>0.006984</td>\n",
-       "      <td>0.005622</td>\n",
-       "      <td>0.007970</td>\n",
-       "      <td>0.005596</td>\n",
-       "      <td>0.007873</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>20</th>\n",
-       "      <td>poet</td>\n",
-       "      <td>0.474255</td>\n",
-       "      <td>-0.005773</td>\n",
-       "      <td>0.048895</td>\n",
-       "      <td>-0.003770</td>\n",
-       "      <td>0.024477</td>\n",
-       "      <td>0.024477</td>\n",
-       "      <td>0.008778</td>\n",
-       "      <td>0.022239</td>\n",
-       "      <td>-0.000177</td>\n",
-       "      <td>0.000052</td>\n",
-       "      <td>-0.000440</td>\n",
-       "      <td>-0.000572</td>\n",
-       "      <td>-0.000004</td>\n",
-       "      <td>-0.000193</td>\n",
-       "      <td>0.000274</td>\n",
-       "      <td>-0.000098</td>\n",
-       "      <td>-0.000419</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>21</th>\n",
-       "      <td>professor</td>\n",
-       "      <td>0.451132</td>\n",
-       "      <td>-0.024563</td>\n",
-       "      <td>-0.013432</td>\n",
-       "      <td>-0.005137</td>\n",
-       "      <td>-0.007428</td>\n",
-       "      <td>-0.019622</td>\n",
-       "      <td>-0.000217</td>\n",
-       "      <td>-0.008027</td>\n",
-       "      <td>-0.013832</td>\n",
-       "      <td>0.003990</td>\n",
-       "      <td>0.003261</td>\n",
-       "      <td>-0.003094</td>\n",
-       "      <td>-0.004693</td>\n",
-       "      <td>0.005113</td>\n",
-       "      <td>-0.001571</td>\n",
-       "      <td>0.002681</td>\n",
-       "      <td>0.000711</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>14</th>\n",
-       "      <td>painter</td>\n",
-       "      <td>0.432000</td>\n",
-       "      <td>0.042427</td>\n",
-       "      <td>0.005912</td>\n",
-       "      <td>-0.003217</td>\n",
-       "      <td>0.011563</td>\n",
-       "      <td>-0.000261</td>\n",
-       "      <td>0.026343</td>\n",
-       "      <td>0.039471</td>\n",
-       "      <td>0.014780</td>\n",
-       "      <td>-0.000525</td>\n",
-       "      <td>-0.001147</td>\n",
-       "      <td>-0.000714</td>\n",
-       "      <td>-0.000242</td>\n",
-       "      <td>-0.000073</td>\n",
-       "      <td>-0.000410</td>\n",
-       "      <td>-0.000619</td>\n",
-       "      <td>-0.001128</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>18</th>\n",
-       "      <td>photographer</td>\n",
-       "      <td>0.366965</td>\n",
-       "      <td>-0.035063</td>\n",
-       "      <td>-0.046535</td>\n",
-       "      <td>-0.027029</td>\n",
-       "      <td>-0.040333</td>\n",
-       "      <td>-0.022692</td>\n",
-       "      <td>-0.063085</td>\n",
-       "      <td>-0.013019</td>\n",
-       "      <td>-0.016972</td>\n",
-       "      <td>-0.000783</td>\n",
-       "      <td>0.002912</td>\n",
-       "      <td>-0.000228</td>\n",
-       "      <td>0.001646</td>\n",
-       "      <td>0.000291</td>\n",
-       "      <td>0.002615</td>\n",
-       "      <td>-0.000942</td>\n",
-       "      <td>-0.000323</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>2</th>\n",
-       "      <td>attorney</td>\n",
-       "      <td>0.359738</td>\n",
-       "      <td>-0.010214</td>\n",
-       "      <td>-0.010573</td>\n",
-       "      <td>-0.006775</td>\n",
-       "      <td>-0.021454</td>\n",
-       "      <td>0.009278</td>\n",
-       "      <td>0.000966</td>\n",
-       "      <td>0.008740</td>\n",
-       "      <td>-0.013176</td>\n",
-       "      <td>0.000730</td>\n",
-       "      <td>0.002510</td>\n",
-       "      <td>0.000961</td>\n",
-       "      <td>0.000600</td>\n",
-       "      <td>-0.000711</td>\n",
-       "      <td>-0.000081</td>\n",
-       "      <td>-0.000800</td>\n",
-       "      <td>-0.000783</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>0</th>\n",
-       "      <td>accountant</td>\n",
-       "      <td>0.355311</td>\n",
-       "      <td>0.015640</td>\n",
-       "      <td>0.007439</td>\n",
-       "      <td>0.066132</td>\n",
-       "      <td>0.027413</td>\n",
-       "      <td>0.035204</td>\n",
-       "      <td>-0.051488</td>\n",
-       "      <td>0.004686</td>\n",
-       "      <td>0.067830</td>\n",
-       "      <td>0.001109</td>\n",
-       "      <td>0.000433</td>\n",
-       "      <td>0.000906</td>\n",
-       "      <td>0.000961</td>\n",
-       "      <td>0.000867</td>\n",
-       "      <td>0.001488</td>\n",
-       "      <td>0.001167</td>\n",
-       "      <td>0.000809</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>6</th>\n",
-       "      <td>dentist</td>\n",
-       "      <td>0.350000</td>\n",
-       "      <td>0.006237</td>\n",
-       "      <td>0.008613</td>\n",
-       "      <td>0.002673</td>\n",
-       "      <td>0.005643</td>\n",
-       "      <td>-0.008910</td>\n",
-       "      <td>-0.007425</td>\n",
-       "      <td>0.020790</td>\n",
-       "      <td>0.001188</td>\n",
-       "      <td>-0.000131</td>\n",
-       "      <td>0.000010</td>\n",
-       "      <td>0.000854</td>\n",
-       "      <td>0.000819</td>\n",
-       "      <td>0.000317</td>\n",
-       "      <td>0.001285</td>\n",
-       "      <td>0.000186</td>\n",
-       "      <td>0.000062</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>9</th>\n",
-       "      <td>filmmaker</td>\n",
-       "      <td>0.342029</td>\n",
-       "      <td>0.091764</td>\n",
-       "      <td>0.002277</td>\n",
-       "      <td>0.048383</td>\n",
-       "      <td>0.040917</td>\n",
-       "      <td>0.028037</td>\n",
-       "      <td>-0.023818</td>\n",
-       "      <td>0.011760</td>\n",
-       "      <td>0.067722</td>\n",
-       "      <td>0.001294</td>\n",
-       "      <td>0.001559</td>\n",
-       "      <td>0.001055</td>\n",
-       "      <td>0.000633</td>\n",
-       "      <td>0.001483</td>\n",
-       "      <td>0.001973</td>\n",
-       "      <td>0.000314</td>\n",
-       "      <td>0.004027</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>3</th>\n",
-       "      <td>chiropractor</td>\n",
-       "      <td>0.285714</td>\n",
-       "      <td>0.072727</td>\n",
-       "      <td>-0.050000</td>\n",
-       "      <td>0.063636</td>\n",
-       "      <td>0.054545</td>\n",
-       "      <td>0.054545</td>\n",
-       "      <td>0.040909</td>\n",
-       "      <td>0.086364</td>\n",
-       "      <td>0.068182</td>\n",
-       "      <td>0.000023</td>\n",
-       "      <td>0.002153</td>\n",
-       "      <td>0.000472</td>\n",
-       "      <td>0.000322</td>\n",
-       "      <td>0.000768</td>\n",
-       "      <td>0.000599</td>\n",
-       "      <td>0.000135</td>\n",
-       "      <td>0.000433</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>16</th>\n",
-       "      <td>pastor</td>\n",
-       "      <td>0.231343</td>\n",
-       "      <td>-0.237081</td>\n",
-       "      <td>-0.282493</td>\n",
-       "      <td>-0.191982</td>\n",
-       "      <td>-0.191669</td>\n",
-       "      <td>-0.205136</td>\n",
-       "      <td>-0.331037</td>\n",
-       "      <td>-0.146571</td>\n",
-       "      <td>-0.133730</td>\n",
-       "      <td>0.000452</td>\n",
-       "      <td>0.001647</td>\n",
-       "      <td>0.001087</td>\n",
-       "      <td>0.000319</td>\n",
-       "      <td>0.000067</td>\n",
-       "      <td>0.001816</td>\n",
-       "      <td>0.000322</td>\n",
-       "      <td>0.000211</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>1</th>\n",
-       "      <td>architect</td>\n",
-       "      <td>0.206704</td>\n",
-       "      <td>-0.017003</td>\n",
-       "      <td>0.084380</td>\n",
-       "      <td>0.016305</td>\n",
-       "      <td>-0.041936</td>\n",
-       "      <td>0.017066</td>\n",
-       "      <td>0.019033</td>\n",
-       "      <td>0.020302</td>\n",
-       "      <td>0.065601</td>\n",
-       "      <td>0.001249</td>\n",
-       "      <td>0.001746</td>\n",
-       "      <td>0.002404</td>\n",
-       "      <td>0.004873</td>\n",
-       "      <td>0.004585</td>\n",
-       "      <td>0.003719</td>\n",
-       "      <td>0.002787</td>\n",
-       "      <td>0.003190</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>4</th>\n",
-       "      <td>comedian</td>\n",
-       "      <td>0.205479</td>\n",
-       "      <td>-0.010920</td>\n",
-       "      <td>-0.071264</td>\n",
-       "      <td>0.012644</td>\n",
-       "      <td>-0.031609</td>\n",
-       "      <td>0.077011</td>\n",
-       "      <td>-0.122989</td>\n",
-       "      <td>-0.047701</td>\n",
-       "      <td>-0.089655</td>\n",
-       "      <td>-0.000036</td>\n",
-       "      <td>0.000040</td>\n",
-       "      <td>0.000302</td>\n",
-       "      <td>0.000079</td>\n",
-       "      <td>-0.000348</td>\n",
-       "      <td>0.000716</td>\n",
-       "      <td>-0.000255</td>\n",
-       "      <td>0.000061</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>24</th>\n",
-       "      <td>software_engineer</td>\n",
-       "      <td>0.168508</td>\n",
-       "      <td>-0.067643</td>\n",
-       "      <td>-0.016230</td>\n",
-       "      <td>-0.048581</td>\n",
-       "      <td>0.022058</td>\n",
-       "      <td>-0.000327</td>\n",
-       "      <td>-0.060345</td>\n",
-       "      <td>0.014542</td>\n",
-       "      <td>0.022929</td>\n",
-       "      <td>0.005779</td>\n",
-       "      <td>0.008481</td>\n",
-       "      <td>0.008296</td>\n",
-       "      <td>0.005068</td>\n",
-       "      <td>0.003385</td>\n",
-       "      <td>0.004794</td>\n",
-       "      <td>0.005113</td>\n",
-       "      <td>0.005113</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>5</th>\n",
-       "      <td>composer</td>\n",
-       "      <td>0.148472</td>\n",
-       "      <td>0.048115</td>\n",
-       "      <td>-0.039065</td>\n",
-       "      <td>-0.016139</td>\n",
-       "      <td>-0.033937</td>\n",
-       "      <td>0.040271</td>\n",
-       "      <td>-0.033937</td>\n",
-       "      <td>0.009502</td>\n",
-       "      <td>0.108296</td>\n",
-       "      <td>0.001531</td>\n",
-       "      <td>0.002686</td>\n",
-       "      <td>0.001599</td>\n",
-       "      <td>0.001950</td>\n",
-       "      <td>0.002317</td>\n",
-       "      <td>0.002515</td>\n",
-       "      <td>0.001599</td>\n",
-       "      <td>0.001351</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>8</th>\n",
-       "      <td>dj</td>\n",
-       "      <td>0.142857</td>\n",
-       "      <td>-0.180556</td>\n",
-       "      <td>-0.250000</td>\n",
-       "      <td>-0.513889</td>\n",
-       "      <td>-0.319444</td>\n",
-       "      <td>-0.263889</td>\n",
-       "      <td>-0.125000</td>\n",
-       "      <td>-0.111111</td>\n",
-       "      <td>-0.347222</td>\n",
-       "      <td>0.000075</td>\n",
-       "      <td>0.001305</td>\n",
-       "      <td>-0.000147</td>\n",
-       "      <td>-0.000129</td>\n",
-       "      <td>0.000150</td>\n",
-       "      <td>0.000225</td>\n",
-       "      <td>-0.000129</td>\n",
-       "      <td>0.000597</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>25</th>\n",
-       "      <td>surgeon</td>\n",
-       "      <td>0.141194</td>\n",
-       "      <td>-0.113909</td>\n",
-       "      <td>-0.238756</td>\n",
-       "      <td>-0.153451</td>\n",
-       "      <td>-0.199074</td>\n",
-       "      <td>-0.057278</td>\n",
-       "      <td>-0.202883</td>\n",
-       "      <td>0.027590</td>\n",
-       "      <td>-0.088485</td>\n",
-       "      <td>0.004425</td>\n",
-       "      <td>0.005501</td>\n",
-       "      <td>0.003698</td>\n",
-       "      <td>0.005469</td>\n",
-       "      <td>0.003234</td>\n",
-       "      <td>0.004943</td>\n",
-       "      <td>0.005466</td>\n",
-       "      <td>0.003292</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>23</th>\n",
-       "      <td>rapper</td>\n",
-       "      <td>0.070423</td>\n",
-       "      <td>0.303030</td>\n",
-       "      <td>-0.224242</td>\n",
-       "      <td>-0.360606</td>\n",
-       "      <td>-0.175758</td>\n",
-       "      <td>-0.160606</td>\n",
-       "      <td>-0.081818</td>\n",
-       "      <td>0.424242</td>\n",
-       "      <td>-0.039394</td>\n",
-       "      <td>0.000354</td>\n",
-       "      <td>-0.000111</td>\n",
-       "      <td>0.000651</td>\n",
-       "      <td>0.000168</td>\n",
-       "      <td>0.000225</td>\n",
-       "      <td>0.000819</td>\n",
-       "      <td>0.000651</td>\n",
-       "      <td>0.000168</td>\n",
-       "    </tr>\n",
-       "  </tbody>\n",
-       "</table>\n",
-       "</div>"
-      ],
-      "text/plain": [
-       "     label_profession  frac_female  debiased_tuned_tpr_gender_gap  \\\n",
-       "13              nurse     0.913333                       0.103781   \n",
-       "7           dietitian     0.900000                       0.015873   \n",
-       "27       yoga_teacher     0.896552                       0.017094   \n",
-       "15          paralegal     0.857143                       0.393939   \n",
-       "12              model     0.819407                       0.555725   \n",
-       "10  interior_designer     0.742424                       0.096038   \n",
-       "22       psychologist     0.615632                      -0.003880   \n",
-       "26            teacher     0.591990                       0.144606   \n",
-       "17   personal_trainer     0.583333                      -0.028571   \n",
-       "11         journalist     0.489489                      -0.008312   \n",
-       "19          physician     0.485414                       0.019000   \n",
-       "20               poet     0.474255                      -0.005773   \n",
-       "21          professor     0.451132                      -0.024563   \n",
-       "14            painter     0.432000                       0.042427   \n",
-       "18       photographer     0.366965                      -0.035063   \n",
-       "2            attorney     0.359738                      -0.010214   \n",
-       "0          accountant     0.355311                       0.015640   \n",
-       "6             dentist     0.350000                       0.006237   \n",
-       "9           filmmaker     0.342029                       0.091764   \n",
-       "3        chiropractor     0.285714                       0.072727   \n",
-       "16             pastor     0.231343                      -0.237081   \n",
-       "1           architect     0.206704                      -0.017003   \n",
-       "4            comedian     0.205479                      -0.010920   \n",
-       "24  software_engineer     0.168508                      -0.067643   \n",
-       "5            composer     0.148472                       0.048115   \n",
-       "8                  dj     0.142857                      -0.180556   \n",
-       "25            surgeon     0.141194                      -0.113909   \n",
-       "23             rapper     0.070423                       0.303030   \n",
-       "\n",
-       "    debiased_untuned_tpr_gender_gap  strong_tuned_tpr_gender_gap  \\\n",
-       "13                         0.094142                     0.103406   \n",
-       "7                          0.259259                     0.052910   \n",
-       "27                        -0.021368                    -0.008547   \n",
-       "15                         0.454545                     0.242424   \n",
-       "12                         0.568883                     0.563826   \n",
-       "10                         0.295318                     0.180072   \n",
-       "22                         0.014645                     0.001366   \n",
-       "26                         0.177934                     0.196176   \n",
-       "17                        -0.007143                     0.042857   \n",
-       "11                         0.053362                     0.009611   \n",
-       "19                         0.018553                     0.032291   \n",
-       "20                         0.048895                    -0.003770   \n",
-       "21                        -0.013432                    -0.005137   \n",
-       "14                         0.005912                    -0.003217   \n",
-       "18                        -0.046535                    -0.027029   \n",
-       "2                         -0.010573                    -0.006775   \n",
-       "0                          0.007439                     0.066132   \n",
-       "6                          0.008613                     0.002673   \n",
-       "9                          0.002277                     0.048383   \n",
-       "3                         -0.050000                     0.063636   \n",
-       "16                        -0.282493                    -0.191982   \n",
-       "1                          0.084380                     0.016305   \n",
-       "4                         -0.071264                     0.012644   \n",
-       "24                        -0.016230                    -0.048581   \n",
-       "5                         -0.039065                    -0.016139   \n",
-       "8                         -0.250000                    -0.513889   \n",
-       "25                        -0.238756                    -0.153451   \n",
-       "23                        -0.224242                    -0.360606   \n",
-       "\n",
-       "    glove_tuned_tpr_gender_gap  strong_untuned_tpr_gender_gap  \\\n",
-       "13                    0.065413                       0.026483   \n",
-       "7                     0.116402                       0.164021   \n",
-       "27                    0.188034                       0.012821   \n",
-       "15                    0.227273                       0.287879   \n",
-       "12                    0.558769                       0.529409   \n",
-       "10                    0.060024                       0.200480   \n",
-       "22                   -0.067831                      -0.011883   \n",
-       "26                    0.200904                       0.122057   \n",
-       "17                    0.078571                       0.014286   \n",
-       "11                    0.062468                       0.033670   \n",
-       "19                    0.041488                       0.037069   \n",
-       "20                    0.024477                       0.024477   \n",
-       "21                   -0.007428                      -0.019622   \n",
-       "14                    0.011563                      -0.000261   \n",
-       "18                   -0.040333                      -0.022692   \n",
-       "2                    -0.021454                       0.009278   \n",
-       "0                     0.027413                       0.035204   \n",
-       "6                     0.005643                      -0.008910   \n",
-       "9                     0.040917                       0.028037   \n",
-       "3                     0.054545                       0.054545   \n",
-       "16                   -0.191669                      -0.205136   \n",
-       "1                    -0.041936                       0.017066   \n",
-       "4                    -0.031609                       0.077011   \n",
-       "24                    0.022058                      -0.000327   \n",
-       "5                    -0.033937                       0.040271   \n",
-       "8                    -0.319444                      -0.263889   \n",
-       "25                   -0.199074                      -0.057278   \n",
-       "23                   -0.175758                      -0.160606   \n",
-       "\n",
-       "    glove_untuned_tpr_gender_gap  scrubbed_tuned_tpr_gender_gap  \\\n",
-       "13                      0.074677                       0.033034   \n",
-       "7                       0.174603                       0.079365   \n",
-       "27                     -0.047009                       0.089744   \n",
-       "15                      0.409091                       0.318182   \n",
-       "12                      0.585330                       0.492488   \n",
-       "10                      0.141657                       0.080432   \n",
-       "22                     -0.019193                      -0.024338   \n",
-       "26                      0.162972                       0.164425   \n",
-       "17                     -0.121429                       0.064286   \n",
-       "11                      0.074991                      -0.001792   \n",
-       "19                      0.002383                       0.063199   \n",
-       "20                      0.008778                       0.022239   \n",
-       "21                     -0.000217                      -0.008027   \n",
-       "14                      0.026343                       0.039471   \n",
-       "18                     -0.063085                      -0.013019   \n",
-       "2                       0.000966                       0.008740   \n",
-       "0                      -0.051488                       0.004686   \n",
-       "6                      -0.007425                       0.020790   \n",
-       "9                      -0.023818                       0.011760   \n",
-       "3                       0.040909                       0.086364   \n",
-       "16                     -0.331037                      -0.146571   \n",
-       "1                       0.019033                       0.020302   \n",
-       "4                      -0.122989                      -0.047701   \n",
-       "24                     -0.060345                       0.014542   \n",
-       "5                      -0.033937                       0.009502   \n",
-       "8                      -0.125000                      -0.111111   \n",
-       "25                     -0.202883                       0.027590   \n",
-       "23                     -0.081818                       0.424242   \n",
-       "\n",
-       "    scrubbed_untuned_tpr_gender_gap  debiased_tuned_tnr_gender_gap  \\\n",
-       "13                         0.056990                      -0.013451   \n",
-       "7                          0.074074                      -0.002258   \n",
-       "27                        -0.021368                      -0.001124   \n",
-       "15                         0.242424                      -0.000208   \n",
-       "12                         0.532944                      -0.000361   \n",
-       "10                         0.100840                       0.000166   \n",
-       "22                         0.019479                      -0.005905   \n",
-       "26                         0.122142                      -0.002170   \n",
-       "17                        -0.021429                      -0.000298   \n",
-       "11                         0.036690                       0.004109   \n",
-       "19                         0.021449                       0.009150   \n",
-       "20                        -0.000177                       0.000052   \n",
-       "21                        -0.013832                       0.003990   \n",
-       "14                         0.014780                      -0.000525   \n",
-       "18                        -0.016972                      -0.000783   \n",
-       "2                         -0.013176                       0.000730   \n",
-       "0                          0.067830                       0.001109   \n",
-       "6                          0.001188                      -0.000131   \n",
-       "9                          0.067722                       0.001294   \n",
-       "3                          0.068182                       0.000023   \n",
-       "16                        -0.133730                       0.000452   \n",
-       "1                          0.065601                       0.001249   \n",
-       "4                         -0.089655                      -0.000036   \n",
-       "24                         0.022929                       0.005779   \n",
-       "5                          0.108296                       0.001531   \n",
-       "8                         -0.347222                       0.000075   \n",
-       "25                        -0.088485                       0.004425   \n",
-       "23                        -0.039394                       0.000354   \n",
-       "\n",
-       "    debiased_untuned_tnr_gender_gap  strong_tuned_tnr_gender_gap  \\\n",
-       "13                        -0.009189                    -0.006368   \n",
-       "7                         -0.001876                    -0.001783   \n",
-       "27                        -0.001105                    -0.001012   \n",
-       "15                        -0.001735                    -0.000004   \n",
-       "12                        -0.001767                    -0.000536   \n",
-       "10                        -0.000577                    -0.000207   \n",
-       "22                        -0.004092                    -0.004249   \n",
-       "26                        -0.002134                    -0.002453   \n",
-       "17                         0.000296                    -0.000205   \n",
-       "11                        -0.000966                     0.001937   \n",
-       "19                         0.007732                     0.007114   \n",
-       "20                        -0.000440                    -0.000572   \n",
-       "21                         0.003261                    -0.003094   \n",
-       "14                        -0.001147                    -0.000714   \n",
-       "18                         0.002912                    -0.000228   \n",
-       "2                          0.002510                     0.000961   \n",
-       "0                          0.000433                     0.000906   \n",
-       "6                          0.000010                     0.000854   \n",
-       "9                          0.001559                     0.001055   \n",
-       "3                          0.002153                     0.000472   \n",
-       "16                         0.001647                     0.001087   \n",
-       "1                          0.001746                     0.002404   \n",
-       "4                          0.000040                     0.000302   \n",
-       "24                         0.008481                     0.008296   \n",
-       "5                          0.002686                     0.001599   \n",
-       "8                          0.001305                    -0.000147   \n",
-       "25                         0.005501                     0.003698   \n",
-       "23                        -0.000111                     0.000651   \n",
-       "\n",
-       "    glove_tuned_tnr_gender_gap  strong_untuned_tnr_gender_gap  \\\n",
-       "13                   -0.009822                      -0.004472   \n",
-       "7                    -0.001969                      -0.001855   \n",
-       "27                   -0.001012                      -0.000900   \n",
-       "15                    0.000185                       0.000054   \n",
-       "12                   -0.001274                       0.001565   \n",
-       "10                    0.000055                      -0.000002   \n",
-       "22                   -0.001270                      -0.006115   \n",
-       "26                   -0.004788                      -0.000893   \n",
-       "17                   -0.000242                      -0.000520   \n",
-       "11                   -0.000169                       0.001959   \n",
-       "19                    0.006984                       0.005622   \n",
-       "20                   -0.000004                      -0.000193   \n",
-       "21                   -0.004693                       0.005113   \n",
-       "14                   -0.000242                      -0.000073   \n",
-       "18                    0.001646                       0.000291   \n",
-       "2                     0.000600                      -0.000711   \n",
-       "0                     0.000961                       0.000867   \n",
-       "6                     0.000819                       0.000317   \n",
-       "9                     0.000633                       0.001483   \n",
-       "3                     0.000322                       0.000768   \n",
-       "16                    0.000319                       0.000067   \n",
-       "1                     0.004873                       0.004585   \n",
-       "4                     0.000079                      -0.000348   \n",
-       "24                    0.005068                       0.003385   \n",
-       "5                     0.001950                       0.002317   \n",
-       "8                    -0.000129                       0.000150   \n",
-       "25                    0.005469                       0.003234   \n",
-       "23                    0.000168                       0.000225   \n",
-       "\n",
-       "    glove_untuned_tnr_gender_gap  scrubbed_tuned_tnr_gender_gap  \\\n",
-       "13                     -0.008299                      -0.008484   \n",
-       "7                      -0.002917                      -0.001989   \n",
-       "27                     -0.001217                      -0.001441   \n",
-       "15                     -0.000748                       0.000073   \n",
-       "12                     -0.000913                      -0.000300   \n",
-       "10                     -0.000988                       0.000313   \n",
-       "22                     -0.005507                      -0.002363   \n",
-       "26                     -0.002372                      -0.003146   \n",
-       "17                     -0.000354                      -0.000854   \n",
-       "11                     -0.000280                       0.003475   \n",
-       "19                      0.007970                       0.005596   \n",
-       "20                      0.000274                      -0.000098   \n",
-       "21                     -0.001571                       0.002681   \n",
-       "14                     -0.000410                      -0.000619   \n",
-       "18                      0.002615                      -0.000942   \n",
-       "2                      -0.000081                      -0.000800   \n",
-       "0                       0.001488                       0.001167   \n",
-       "6                       0.001285                       0.000186   \n",
-       "9                       0.001973                       0.000314   \n",
-       "3                       0.000599                       0.000135   \n",
-       "16                      0.001816                       0.000322   \n",
-       "1                       0.003719                       0.002787   \n",
-       "4                       0.000716                      -0.000255   \n",
-       "24                      0.004794                       0.005113   \n",
-       "5                       0.002515                       0.001599   \n",
-       "8                       0.000225                      -0.000129   \n",
-       "25                      0.004943                       0.005466   \n",
-       "23                      0.000819                       0.000651   \n",
-       "\n",
-       "    scrubbed_untuned_tnr_gender_gap  \n",
-       "13                        -0.007496  \n",
-       "7                         -0.001423  \n",
-       "27                        -0.000808  \n",
-       "15                        -0.000170  \n",
-       "12                         0.000077  \n",
-       "10                         0.000090  \n",
-       "22                        -0.004891  \n",
-       "26                        -0.000795  \n",
-       "17                        -0.000316  \n",
-       "11                         0.000804  \n",
-       "19                         0.007873  \n",
-       "20                        -0.000419  \n",
-       "21                         0.000711  \n",
-       "14                        -0.001128  \n",
-       "18                        -0.000323  \n",
-       "2                         -0.000783  \n",
-       "0                          0.000809  \n",
-       "6                          0.000062  \n",
-       "9                          0.004027  \n",
-       "3                          0.000433  \n",
-       "16                         0.000211  \n",
-       "1                          0.003190  \n",
-       "4                          0.000061  \n",
-       "24                         0.005113  \n",
-       "5                          0.001351  \n",
-       "8                          0.000597  \n",
-       "25                         0.003292  \n",
-       "23                         0.000168  "
-      ]
-     },
-     "execution_count": 29,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
     "gender_gap_df.sort_values('frac_female', ascending = False)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 30,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1765,41 +446,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 31,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "debiased_tuned\n",
-      "0.6071428571428571\n",
-      "debiased_untuned\n",
-      "0.42857142857142855\n",
-      "strong_tuned\n",
-      "0.6071428571428571\n",
-      "glove_tuned\n",
-      "0.5714285714285714\n",
-      "strong_untuned\n",
-      "0.6071428571428571\n",
-      "glove_untuned\n",
-      "1.0\n",
-      "scrubbed_tuned\n",
-      "0.5714285714285714\n",
-      "scrubbed_untuned\n",
-      "0.6428571428571429\n"
-     ]
-    }
-   ],
+   "execution_count": null,
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [],
    "source": [
-    "for _model in MODEL_NAMES.values():\n",
-    "    print(_model)\n",
-    "    print(compute_fraction_improved(gender_gap_df, 'glove_untuned_tpr_gender_gap', '{}_tpr_gender_gap'.format(_model)))"
+    "# for _model in MODEL_NAMES.values():\n",
+    "#     print(_model)\n",
+    "#     print(compute_fraction_improved(gender_gap_df, 'glove_untuned_tpr_gender_gap', '{}_tpr_gender_gap'.format(_model)))"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 32,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1810,79 +470,25 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 33,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "debiased_tuned_tpr_gender_gap      0.025907\n",
-       "debiased_untuned_tpr_gender_gap    0.035721\n",
-       "strong_tuned_tpr_gender_gap        0.033332\n",
-       "glove_tuned_tpr_gender_gap         0.024980\n",
-       "strong_untuned_tpr_gender_gap      0.021577\n",
-       "glove_untuned_tpr_gender_gap       0.029238\n",
-       "scrubbed_tuned_tpr_gender_gap      0.022480\n",
-       "scrubbed_untuned_tpr_gender_gap    0.020245\n",
-       "debiased_tuned_tnr_gender_gap      0.000014\n",
-       "debiased_untuned_tnr_gender_gap    0.000012\n",
-       "strong_tuned_tnr_gender_gap        0.000008\n",
-       "glove_tuned_tnr_gender_gap         0.000010\n",
-       "strong_untuned_tnr_gender_gap      0.000006\n",
-       "glove_untuned_tnr_gender_gap       0.000010\n",
-       "scrubbed_tuned_tnr_gender_gap      0.000008\n",
-       "scrubbed_untuned_tnr_gender_gap    0.000008\n",
-       "dtype: float64"
-      ]
-     },
-     "execution_count": 33,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
     "gender_gap_df[gender_gap_cols].apply(lambda x: np.mean(x**2))"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 34,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "debiased_tuned_tpr_gender_gap      0.095339\n",
-       "debiased_untuned_tpr_gender_gap    0.120188\n",
-       "strong_tuned_tpr_gender_gap        0.106208\n",
-       "glove_tuned_tpr_gender_gap         0.104158\n",
-       "strong_untuned_tpr_gender_gap      0.088024\n",
-       "glove_untuned_tpr_gender_gap       0.107622\n",
-       "scrubbed_tuned_tpr_gender_gap      0.086712\n",
-       "scrubbed_untuned_tpr_gender_gap    0.086036\n",
-       "debiased_tuned_tnr_gender_gap      0.002205\n",
-       "debiased_untuned_tnr_gender_gap    0.002407\n",
-       "strong_tuned_tnr_gender_gap        0.001889\n",
-       "glove_tuned_tnr_gender_gap         0.001991\n",
-       "strong_untuned_tnr_gender_gap      0.001717\n",
-       "glove_untuned_tnr_gender_gap       0.002193\n",
-       "scrubbed_tuned_tnr_gender_gap      0.001832\n",
-       "scrubbed_untuned_tnr_gender_gap    0.001694\n",
-       "dtype: float64"
-      ]
-     },
-     "execution_count": 34,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
     "gender_gap_df[gender_gap_cols].apply(lambda x: np.mean(np.abs(x)))"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 35,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1900,58 +506,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 36,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<matplotlib.figure.Figure at 0x1035efc90>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<matplotlib.figure.Figure at 0x114a41450>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<matplotlib.figure.Figure at 0x1054547d0>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<matplotlib.figure.Figure at 0x105461bd0>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
+   "outputs": [],
    "source": [
     "for _model in MODEL_NAMES.values():\n",
     "    if 'untuned' in _model:\n",
@@ -1960,92 +517,9 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 37,
+   "execution_count": null,
    "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/html": [
-       "<div>\n",
-       "<style scoped>\n",
-       "    .dataframe tbody tr th:only-of-type {\n",
-       "        vertical-align: middle;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe tbody tr th {\n",
-       "        vertical-align: top;\n",
-       "    }\n",
-       "\n",
-       "    .dataframe thead th {\n",
-       "        text-align: right;\n",
-       "    }\n",
-       "</style>\n",
-       "<table border=\"1\" class=\"dataframe\">\n",
-       "  <thead>\n",
-       "    <tr style=\"text-align: right;\">\n",
-       "      <th></th>\n",
-       "      <th>frac_female</th>\n",
-       "    </tr>\n",
-       "  </thead>\n",
-       "  <tbody>\n",
-       "    <tr>\n",
-       "      <th>frac_female</th>\n",
-       "      <td>1.000000</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>debiased_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.442007</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>debiased_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.738924</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>strong_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.645211</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>glove_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.711791</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>strong_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.627283</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>glove_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.647357</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>scrubbed_tuned_tpr_gender_gap</th>\n",
-       "      <td>0.341492</td>\n",
-       "    </tr>\n",
-       "    <tr>\n",
-       "      <th>scrubbed_untuned_tpr_gender_gap</th>\n",
-       "      <td>0.520765</td>\n",
-       "    </tr>\n",
-       "  </tbody>\n",
-       "</table>\n",
-       "</div>"
-      ],
-      "text/plain": [
-       "                                 frac_female\n",
-       "frac_female                         1.000000\n",
-       "debiased_tuned_tpr_gender_gap       0.442007\n",
-       "debiased_untuned_tpr_gender_gap     0.738924\n",
-       "strong_tuned_tpr_gender_gap         0.645211\n",
-       "glove_tuned_tpr_gender_gap          0.711791\n",
-       "strong_untuned_tpr_gender_gap       0.627283\n",
-       "glove_untuned_tpr_gender_gap        0.647357\n",
-       "scrubbed_tuned_tpr_gender_gap       0.341492\n",
-       "scrubbed_untuned_tpr_gender_gap     0.520765"
-      ]
-     },
-     "execution_count": 37,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
+   "outputs": [],
    "source": [
     "results_df[['frac_female']+['{}_tpr_gender_gap'.format(_model) for _model in MODEL_NAMES.values()]].corr()[['frac_female']]"
    ]
@@ -2060,9 +534,9 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "models_eval_py2",
+   "display_name": "Python 2",
    "language": "python",
-   "name": "models_eval_py2"
+   "name": "python2"
   },
   "language_info": {
    "codemirror_mode": {
@@ -2074,7 +548,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython2",
-   "version": "2.7.10"
+   "version": "2.7.14+"
   }
  },
  "nbformat": 4,