jbloom's picture
Upload folder using huggingface_hub
ceef8ed verified
raw
history blame contribute delete
No virus
30.3 kB
<style type="text/css">
#T_53d0e_row0_col2, #T_53d0e_row0_col3, #T_53d0e_row0_col5, #T_53d0e_row0_col6, #T_53d0e_row0_col10, #T_53d0e_row0_col11, #T_53d0e_row0_col16, #T_53d0e_row0_col17, #T_53d0e_row1_col6, #T_53d0e_row1_col17, #T_53d0e_row2_col6, #T_53d0e_row2_col17, #T_53d0e_row3_col6, #T_53d0e_row3_col17, #T_53d0e_row4_col6, #T_53d0e_row4_col13, #T_53d0e_row4_col17, #T_53d0e_row5_col6, #T_53d0e_row5_col17, #T_53d0e_row6_col6, #T_53d0e_row6_col17, #T_53d0e_row7_col6, #T_53d0e_row7_col17, #T_53d0e_row8_col6, #T_53d0e_row8_col17, #T_53d0e_row9_col4, #T_53d0e_row9_col6, #T_53d0e_row9_col7, #T_53d0e_row9_col8, #T_53d0e_row9_col9, #T_53d0e_row9_col12, #T_53d0e_row9_col17, #T_53d0e_row10_col6, #T_53d0e_row10_col17, #T_53d0e_row11_col6, #T_53d0e_row11_col14, #T_53d0e_row11_col15, #T_53d0e_row11_col17 {
background-color: #440154;
color: #f1f1f1;
}
#T_53d0e_row0_col4 {
background-color: #31b57b;
color: #f1f1f1;
}
#T_53d0e_row0_col7 {
background-color: #2eb37c;
color: #f1f1f1;
}
#T_53d0e_row0_col8, #T_53d0e_row0_col9, #T_53d0e_row0_col12, #T_53d0e_row0_col15, #T_53d0e_row1_col4, #T_53d0e_row1_col7, #T_53d0e_row1_col9, #T_53d0e_row1_col14, #T_53d0e_row11_col2, #T_53d0e_row11_col3, #T_53d0e_row11_col5, #T_53d0e_row11_col10, #T_53d0e_row11_col11, #T_53d0e_row11_col13, #T_53d0e_row11_col16 {
background-color: #fde725;
color: #000000;
}
#T_53d0e_row0_col13 {
background-color: #28ae80;
color: #f1f1f1;
}
#T_53d0e_row0_col14, #T_53d0e_row8_col9 {
background-color: #26828e;
color: #f1f1f1;
}
#T_53d0e_row1_col2 {
background-color: #482173;
color: #f1f1f1;
}
#T_53d0e_row1_col3 {
background-color: #46085c;
color: #f1f1f1;
}
#T_53d0e_row1_col5 {
background-color: #46075a;
color: #f1f1f1;
}
#T_53d0e_row1_col8 {
background-color: #fbe723;
color: #000000;
}
#T_53d0e_row1_col10, #T_53d0e_row1_col11, #T_53d0e_row10_col15 {
background-color: #481a6c;
color: #f1f1f1;
}
#T_53d0e_row1_col12, #T_53d0e_row10_col13 {
background-color: #e7e419;
color: #000000;
}
#T_53d0e_row1_col13 {
background-color: #f4e61e;
color: #000000;
}
#T_53d0e_row1_col15, #T_53d0e_row2_col9 {
background-color: #e5e419;
color: #000000;
}
#T_53d0e_row1_col16 {
background-color: #450457;
color: #f1f1f1;
}
#T_53d0e_row2_col2, #T_53d0e_row5_col13 {
background-color: #433e85;
color: #f1f1f1;
}
#T_53d0e_row2_col3, #T_53d0e_row5_col16 {
background-color: #481c6e;
color: #f1f1f1;
}
#T_53d0e_row2_col4, #T_53d0e_row10_col5 {
background-color: #3dbc74;
color: #f1f1f1;
}
#T_53d0e_row2_col5, #T_53d0e_row10_col4, #T_53d0e_row10_col7, #T_53d0e_row10_col12 {
background-color: #481b6d;
color: #f1f1f1;
}
#T_53d0e_row2_col7 {
background-color: #40bd72;
color: #f1f1f1;
}
#T_53d0e_row2_col8 {
background-color: #e2e418;
color: #000000;
}
#T_53d0e_row2_col10, #T_53d0e_row2_col11 {
background-color: #482475;
color: #f1f1f1;
}
#T_53d0e_row2_col12 {
background-color: #b5de2b;
color: #000000;
}
#T_53d0e_row2_col13 {
background-color: #f6e620;
color: #000000;
}
#T_53d0e_row2_col14 {
background-color: #65cb5e;
color: #000000;
}
#T_53d0e_row2_col15 {
background-color: #d2e21b;
color: #000000;
}
#T_53d0e_row2_col16 {
background-color: #450559;
color: #f1f1f1;
}
#T_53d0e_row3_col2 {
background-color: #38588c;
color: #f1f1f1;
}
#T_53d0e_row3_col3, #T_53d0e_row6_col7 {
background-color: #46337f;
color: #f1f1f1;
}
#T_53d0e_row3_col4 {
background-color: #26818e;
color: #f1f1f1;
}
#T_53d0e_row3_col5 {
background-color: #463480;
color: #f1f1f1;
}
#T_53d0e_row3_col7, #T_53d0e_row11_col9 {
background-color: #27808e;
color: #f1f1f1;
}
#T_53d0e_row3_col8, #T_53d0e_row3_col9 {
background-color: #b2dd2d;
color: #000000;
}
#T_53d0e_row3_col10, #T_53d0e_row3_col11 {
background-color: #414287;
color: #f1f1f1;
}
#T_53d0e_row3_col12 {
background-color: #60ca60;
color: #000000;
}
#T_53d0e_row3_col13 {
background-color: #37b878;
color: #f1f1f1;
}
#T_53d0e_row3_col14, #T_53d0e_row9_col2 {
background-color: #86d549;
color: #000000;
}
#T_53d0e_row3_col15 {
background-color: #f8e621;
color: #000000;
}
#T_53d0e_row3_col16 {
background-color: #460a5d;
color: #f1f1f1;
}
#T_53d0e_row4_col2 {
background-color: #2d708e;
color: #f1f1f1;
}
#T_53d0e_row4_col3, #T_53d0e_row6_col12 {
background-color: #3b528b;
color: #f1f1f1;
}
#T_53d0e_row4_col4 {
background-color: #50c46a;
color: #000000;
}
#T_53d0e_row4_col5 {
background-color: #365c8d;
color: #f1f1f1;
}
#T_53d0e_row4_col7, #T_53d0e_row8_col2 {
background-color: #52c569;
color: #000000;
}
#T_53d0e_row4_col8 {
background-color: #a0da39;
color: #000000;
}
#T_53d0e_row4_col9 {
background-color: #95d840;
color: #000000;
}
#T_53d0e_row4_col10, #T_53d0e_row4_col11 {
background-color: #3f4889;
color: #f1f1f1;
}
#T_53d0e_row4_col12 {
background-color: #1fa187;
color: #f1f1f1;
}
#T_53d0e_row4_col14 {
background-color: #addc30;
color: #000000;
}
#T_53d0e_row4_col15 {
background-color: #b0dd2f;
color: #000000;
}
#T_53d0e_row4_col16 {
background-color: #471365;
color: #f1f1f1;
}
#T_53d0e_row5_col2 {
background-color: #25858e;
color: #f1f1f1;
}
#T_53d0e_row5_col3 {
background-color: #31668e;
color: #f1f1f1;
}
#T_53d0e_row5_col4 {
background-color: #1fa088;
color: #f1f1f1;
}
#T_53d0e_row5_col5 {
background-color: #33638d;
color: #f1f1f1;
}
#T_53d0e_row5_col7, #T_53d0e_row10_col14 {
background-color: #1e9d89;
color: #f1f1f1;
}
#T_53d0e_row5_col8 {
background-color: #70cf57;
color: #000000;
}
#T_53d0e_row5_col9 {
background-color: #77d153;
color: #000000;
}
#T_53d0e_row5_col10 {
background-color: #3c4f8a;
color: #f1f1f1;
}
#T_53d0e_row5_col11, #T_53d0e_row8_col16, #T_53d0e_row9_col15 {
background-color: #3d4e8a;
color: #f1f1f1;
}
#T_53d0e_row5_col12, #T_53d0e_row8_col15, #T_53d0e_row9_col10 {
background-color: #297a8e;
color: #f1f1f1;
}
#T_53d0e_row5_col14, #T_53d0e_row6_col14 {
background-color: #a5db36;
color: #000000;
}
#T_53d0e_row5_col15 {
background-color: #69cd5b;
color: #000000;
}
#T_53d0e_row6_col2 {
background-color: #1e9b8a;
color: #f1f1f1;
}
#T_53d0e_row6_col3, #T_53d0e_row11_col8 {
background-color: #277f8e;
color: #f1f1f1;
}
#T_53d0e_row6_col4 {
background-color: #46327e;
color: #f1f1f1;
}
#T_53d0e_row6_col5, #T_53d0e_row9_col11 {
background-color: #2a788e;
color: #f1f1f1;
}
#T_53d0e_row6_col8 {
background-color: #1f958b;
color: #f1f1f1;
}
#T_53d0e_row6_col9 {
background-color: #1f9f88;
color: #f1f1f1;
}
#T_53d0e_row6_col10, #T_53d0e_row6_col11 {
background-color: #39558c;
color: #f1f1f1;
}
#T_53d0e_row6_col13 {
background-color: #cde11d;
color: #000000;
}
#T_53d0e_row6_col15 {
background-color: #32b67a;
color: #f1f1f1;
}
#T_53d0e_row6_col16 {
background-color: #482878;
color: #f1f1f1;
}
#T_53d0e_row7_col2 {
background-color: #2ab07f;
color: #f1f1f1;
}
#T_53d0e_row7_col3 {
background-color: #1f968b;
color: #f1f1f1;
}
#T_53d0e_row7_col4 {
background-color: #2c738e;
color: #f1f1f1;
}
#T_53d0e_row7_col5 {
background-color: #21918c;
color: #f1f1f1;
}
#T_53d0e_row7_col7 {
background-color: #2b748e;
color: #f1f1f1;
}
#T_53d0e_row7_col8 {
background-color: #20a386;
color: #f1f1f1;
}
#T_53d0e_row7_col9 {
background-color: #24aa83;
color: #f1f1f1;
}
#T_53d0e_row7_col10 {
background-color: #355f8d;
color: #f1f1f1;
}
#T_53d0e_row7_col11 {
background-color: #355e8d;
color: #f1f1f1;
}
#T_53d0e_row7_col12 {
background-color: #46307e;
color: #f1f1f1;
}
#T_53d0e_row7_col13, #T_53d0e_row9_col13 {
background-color: #dae319;
color: #000000;
}
#T_53d0e_row7_col14 {
background-color: #9bd93c;
color: #000000;
}
#T_53d0e_row7_col15 {
background-color: #1e9c89;
color: #f1f1f1;
}
#T_53d0e_row7_col16 {
background-color: #453781;
color: #f1f1f1;
}
#T_53d0e_row8_col3 {
background-color: #2db27d;
color: #f1f1f1;
}
#T_53d0e_row8_col4, #T_53d0e_row8_col7 {
background-color: #3d4d8a;
color: #f1f1f1;
}
#T_53d0e_row8_col5 {
background-color: #21a585;
color: #f1f1f1;
}
#T_53d0e_row8_col8 {
background-color: #2c728e;
color: #f1f1f1;
}
#T_53d0e_row8_col10 {
background-color: #306a8e;
color: #f1f1f1;
}
#T_53d0e_row8_col11 {
background-color: #30698e;
color: #f1f1f1;
}
#T_53d0e_row8_col12 {
background-color: #481668;
color: #f1f1f1;
}
#T_53d0e_row8_col13 {
background-color: #eae51a;
color: #000000;
}
#T_53d0e_row8_col14 {
background-color: #67cc5c;
color: #000000;
}
#T_53d0e_row9_col3 {
background-color: #44bf70;
color: #f1f1f1;
}
#T_53d0e_row9_col5 {
background-color: #4ac16d;
color: #000000;
}
#T_53d0e_row9_col14 {
background-color: #42be71;
color: #f1f1f1;
}
#T_53d0e_row9_col16 {
background-color: #2e6f8e;
color: #f1f1f1;
}
#T_53d0e_row10_col2 {
background-color: #c2df23;
color: #000000;
}
#T_53d0e_row10_col3 {
background-color: #6ccd5a;
color: #000000;
}
#T_53d0e_row10_col8 {
background-color: #471063;
color: #f1f1f1;
}
#T_53d0e_row10_col9 {
background-color: #453882;
color: #f1f1f1;
}
#T_53d0e_row10_col10 {
background-color: #1f978b;
color: #f1f1f1;
}
#T_53d0e_row10_col11 {
background-color: #1f948c;
color: #f1f1f1;
}
#T_53d0e_row10_col16 {
background-color: #20a486;
color: #f1f1f1;
}
#T_53d0e_row11_col4 {
background-color: #48c16e;
color: #f1f1f1;
}
#T_53d0e_row11_col7 {
background-color: #46c06f;
color: #f1f1f1;
}
#T_53d0e_row11_col12 {
background-color: #5cc863;
color: #000000;
}
</style>
<table id="T_53d0e">
<thead>
<tr>
<th class="blank level0" >&nbsp;</th>
<th id="T_53d0e_level0_col0" class="col_heading level0 col0" >version</th>
<th id="T_53d0e_level0_col1" class="col_heading level0 col1" >d_sae</th>
<th id="T_53d0e_level0_col2" class="col_heading level0 col2" >layer</th>
<th id="T_53d0e_level0_col3" class="col_heading level0 col3" >kl_div_with_sae</th>
<th id="T_53d0e_level0_col4" class="col_heading level0 col4" >kl_div_with_ablation</th>
<th id="T_53d0e_level0_col5" class="col_heading level0 col5" >ce_loss_with_sae</th>
<th id="T_53d0e_level0_col6" class="col_heading level0 col6" >ce_loss_without_sae</th>
<th id="T_53d0e_level0_col7" class="col_heading level0 col7" >ce_loss_with_ablation</th>
<th id="T_53d0e_level0_col8" class="col_heading level0 col8" >kl_div_score</th>
<th id="T_53d0e_level0_col9" class="col_heading level0 col9" >ce_loss_score</th>
<th id="T_53d0e_level0_col10" class="col_heading level0 col10" >l2_norm_in</th>
<th id="T_53d0e_level0_col11" class="col_heading level0 col11" >l2_norm_out</th>
<th id="T_53d0e_level0_col12" class="col_heading level0 col12" >l2_ratio</th>
<th id="T_53d0e_level0_col13" class="col_heading level0 col13" >l0</th>
<th id="T_53d0e_level0_col14" class="col_heading level0 col14" >l1</th>
<th id="T_53d0e_level0_col15" class="col_heading level0 col15" >explained_variance</th>
<th id="T_53d0e_level0_col16" class="col_heading level0 col16" >mse</th>
<th id="T_53d0e_level0_col17" class="col_heading level0 col17" >total_tokens_evaluated</th>
<th id="T_53d0e_level0_col18" class="col_heading level0 col18" >filepath</th>
</tr>
</thead>
<tbody>
<tr>
<th id="T_53d0e_level0_row0" class="row_heading level0 row0" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_0/metrics.json</th>
<td id="T_53d0e_row0_col0" class="data row0 col0" >5</td>
<td id="T_53d0e_row0_col1" class="data row0 col1" >128</td>
<td id="T_53d0e_row0_col2" class="data row0 col2" >0</td>
<td id="T_53d0e_row0_col3" class="data row0 col3" >0.003843</td>
<td id="T_53d0e_row0_col4" class="data row0 col4" >12.480284</td>
<td id="T_53d0e_row0_col5" class="data row0 col5" >3.603422</td>
<td id="T_53d0e_row0_col6" class="data row0 col6" >3.599065</td>
<td id="T_53d0e_row0_col7" class="data row0 col7" >15.861977</td>
<td id="T_53d0e_row0_col8" class="data row0 col8" >0.999692</td>
<td id="T_53d0e_row0_col9" class="data row0 col9" >0.999645</td>
<td id="T_53d0e_row0_col10" class="data row0 col10" >32.707962</td>
<td id="T_53d0e_row0_col11" class="data row0 col11" >32.607395</td>
<td id="T_53d0e_row0_col12" class="data row0 col12" >0.996936</td>
<td id="T_53d0e_row0_col13" class="data row0 col13" >31.980795</td>
<td id="T_53d0e_row0_col14" class="data row0 col14" >44.247345</td>
<td id="T_53d0e_row0_col15" class="data row0 col15" >0.976495</td>
<td id="T_53d0e_row0_col16" class="data row0 col16" >5.842685</td>
<td id="T_53d0e_row0_col17" class="data row0 col17" >6144.000000</td>
<td id="T_53d0e_row0_col18" class="data row0 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_0/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row1" class="row_heading level0 row1" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_1/metrics.json</th>
<td id="T_53d0e_row1_col0" class="data row1 col0" >5</td>
<td id="T_53d0e_row1_col1" class="data row1 col1" >128</td>
<td id="T_53d0e_row1_col2" class="data row1 col2" >1</td>
<td id="T_53d0e_row1_col3" class="data row1 col3" >0.006731</td>
<td id="T_53d0e_row1_col4" class="data row1 col4" >16.217104</td>
<td id="T_53d0e_row1_col5" class="data row1 col5" >3.605462</td>
<td id="T_53d0e_row1_col6" class="data row1 col6" >3.599065</td>
<td id="T_53d0e_row1_col7" class="data row1 col7" >19.600266</td>
<td id="T_53d0e_row1_col8" class="data row1 col8" >0.999585</td>
<td id="T_53d0e_row1_col9" class="data row1 col9" >0.999600</td>
<td id="T_53d0e_row1_col10" class="data row1 col10" >56.929867</td>
<td id="T_53d0e_row1_col11" class="data row1 col11" >56.693493</td>
<td id="T_53d0e_row1_col12" class="data row1 col12" >0.995869</td>
<td id="T_53d0e_row1_col13" class="data row1 col13" >31.999023</td>
<td id="T_53d0e_row1_col14" class="data row1 col14" >59.953354</td>
<td id="T_53d0e_row1_col15" class="data row1 col15" >0.972269</td>
<td id="T_53d0e_row1_col16" class="data row1 col16" >31.145605</td>
<td id="T_53d0e_row1_col17" class="data row1 col17" >6144.000000</td>
<td id="T_53d0e_row1_col18" class="data row1 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_1/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row2" class="row_heading level0 row2" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_2/metrics.json</th>
<td id="T_53d0e_row2_col0" class="data row2 col0" >5</td>
<td id="T_53d0e_row2_col1" class="data row2 col1" >128</td>
<td id="T_53d0e_row2_col2" class="data row2 col2" >2</td>
<td id="T_53d0e_row2_col3" class="data row2 col3" >0.014035</td>
<td id="T_53d0e_row2_col4" class="data row2 col4" >12.813511</td>
<td id="T_53d0e_row2_col5" class="data row2 col5" >3.612803</td>
<td id="T_53d0e_row2_col6" class="data row2 col6" >3.599065</td>
<td id="T_53d0e_row2_col7" class="data row2 col7" >16.327873</td>
<td id="T_53d0e_row2_col8" class="data row2 col8" >0.998905</td>
<td id="T_53d0e_row2_col9" class="data row2 col9" >0.998921</td>
<td id="T_53d0e_row2_col10" class="data row2 col10" >68.907532</td>
<td id="T_53d0e_row2_col11" class="data row2 col11" >68.518921</td>
<td id="T_53d0e_row2_col12" class="data row2 col12" >0.993602</td>
<td id="T_53d0e_row2_col13" class="data row2 col13" >31.999350</td>
<td id="T_53d0e_row2_col14" class="data row2 col14" >53.266327</td>
<td id="T_53d0e_row2_col15" class="data row2 col15" >0.969186</td>
<td id="T_53d0e_row2_col16" class="data row2 col16" >49.732628</td>
<td id="T_53d0e_row2_col17" class="data row2 col17" >6144.000000</td>
<td id="T_53d0e_row2_col18" class="data row2 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_2/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row3" class="row_heading level0 row3" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_3/metrics.json</th>
<td id="T_53d0e_row3_col0" class="data row3 col0" >5</td>
<td id="T_53d0e_row3_col1" class="data row3 col1" >128</td>
<td id="T_53d0e_row3_col2" class="data row3 col2" >3</td>
<td id="T_53d0e_row3_col3" class="data row3 col3" >0.023511</td>
<td id="T_53d0e_row3_col4" class="data row3 col4" >10.101868</td>
<td id="T_53d0e_row3_col5" class="data row3 col5" >3.622775</td>
<td id="T_53d0e_row3_col6" class="data row3 col6" >3.599065</td>
<td id="T_53d0e_row3_col7" class="data row3 col7" >13.548822</td>
<td id="T_53d0e_row3_col8" class="data row3 col8" >0.997673</td>
<td id="T_53d0e_row3_col9" class="data row3 col9" >0.997617</td>
<td id="T_53d0e_row3_col10" class="data row3 col10" >103.711441</td>
<td id="T_53d0e_row3_col11" class="data row3 col11" >103.026962</td>
<td id="T_53d0e_row3_col12" class="data row3 col12" >0.989538</td>
<td id="T_53d0e_row3_col13" class="data row3 col13" >31.983074</td>
<td id="T_53d0e_row3_col14" class="data row3 col14" >54.825603</td>
<td id="T_53d0e_row3_col15" class="data row3 col15" >0.975295</td>
<td id="T_53d0e_row3_col16" class="data row3 col16" >89.620079</td>
<td id="T_53d0e_row3_col17" class="data row3 col17" >6144.000000</td>
<td id="T_53d0e_row3_col18" class="data row3 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_3/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row4" class="row_heading level0 row4" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_4/metrics.json</th>
<td id="T_53d0e_row4_col0" class="data row4 col0" >5</td>
<td id="T_53d0e_row4_col1" class="data row4 col1" >128</td>
<td id="T_53d0e_row4_col2" class="data row4 col2" >4</td>
<td id="T_53d0e_row4_col3" class="data row4 col3" >0.037530</td>
<td id="T_53d0e_row4_col4" class="data row4 col4" >13.249713</td>
<td id="T_53d0e_row4_col5" class="data row4 col5" >3.640705</td>
<td id="T_53d0e_row4_col6" class="data row4 col6" >3.599065</td>
<td id="T_53d0e_row4_col7" class="data row4 col7" >16.699104</td>
<td id="T_53d0e_row4_col8" class="data row4 col8" >0.997167</td>
<td id="T_53d0e_row4_col9" class="data row4 col9" >0.996821</td>
<td id="T_53d0e_row4_col10" class="data row4 col10" >111.403282</td>
<td id="T_53d0e_row4_col11" class="data row4 col11" >110.286774</td>
<td id="T_53d0e_row4_col12" class="data row4 col12" >0.984007</td>
<td id="T_53d0e_row4_col13" class="data row4 col13" >31.948568</td>
<td id="T_53d0e_row4_col14" class="data row4 col14" >56.497829</td>
<td id="T_53d0e_row4_col15" class="data row4 col15" >0.963923</td>
<td id="T_53d0e_row4_col16" class="data row4 col16" >153.826294</td>
<td id="T_53d0e_row4_col17" class="data row4 col17" >6144.000000</td>
<td id="T_53d0e_row4_col18" class="data row4 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_4/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row5" class="row_heading level0 row5" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_5/metrics.json</th>
<td id="T_53d0e_row5_col0" class="data row5 col0" >5</td>
<td id="T_53d0e_row5_col1" class="data row5 col1" >128</td>
<td id="T_53d0e_row5_col2" class="data row5 col2" >5</td>
<td id="T_53d0e_row5_col3" class="data row5 col3" >0.047411</td>
<td id="T_53d0e_row5_col4" class="data row5 col4" >11.519682</td>
<td id="T_53d0e_row5_col5" class="data row5 col5" >3.644166</td>
<td id="T_53d0e_row5_col6" class="data row5 col6" >3.599065</td>
<td id="T_53d0e_row5_col7" class="data row5 col7" >14.860109</td>
<td id="T_53d0e_row5_col8" class="data row5 col8" >0.995884</td>
<td id="T_53d0e_row5_col9" class="data row5 col9" >0.995995</td>
<td id="T_53d0e_row5_col10" class="data row5 col10" >119.651489</td>
<td id="T_53d0e_row5_col11" class="data row5 col11" >118.053459</td>
<td id="T_53d0e_row5_col12" class="data row5 col12" >0.978980</td>
<td id="T_53d0e_row5_col13" class="data row5 col13" >31.958008</td>
<td id="T_53d0e_row5_col14" class="data row5 col14" >56.149471</td>
<td id="T_53d0e_row5_col15" class="data row5 col15" >0.952276</td>
<td id="T_53d0e_row5_col16" class="data row5 col16" >238.126266</td>
<td id="T_53d0e_row5_col17" class="data row5 col17" >6144.000000</td>
<td id="T_53d0e_row5_col18" class="data row5 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_5/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row6" class="row_heading level0 row6" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_6/metrics.json</th>
<td id="T_53d0e_row6_col0" class="data row6 col0" >5</td>
<td id="T_53d0e_row6_col1" class="data row6 col1" >128</td>
<td id="T_53d0e_row6_col2" class="data row6 col2" >6</td>
<td id="T_53d0e_row6_col3" class="data row6 col3" >0.060794</td>
<td id="T_53d0e_row6_col4" class="data row6 col4" >6.933250</td>
<td id="T_53d0e_row6_col5" class="data row6 col5" >3.655243</td>
<td id="T_53d0e_row6_col6" class="data row6 col6" >3.599065</td>
<td id="T_53d0e_row6_col7" class="data row6 col7" >10.522690</td>
<td id="T_53d0e_row6_col8" class="data row6 col8" >0.991232</td>
<td id="T_53d0e_row6_col9" class="data row6 col9" >0.991886</td>
<td id="T_53d0e_row6_col10" class="data row6 col10" >128.847931</td>
<td id="T_53d0e_row6_col11" class="data row6 col11" >126.688171</td>
<td id="T_53d0e_row6_col12" class="data row6 col12" >0.974231</td>
<td id="T_53d0e_row6_col13" class="data row6 col13" >31.996094</td>
<td id="T_53d0e_row6_col14" class="data row6 col14" >56.119469</td>
<td id="T_53d0e_row6_col15" class="data row6 col15" >0.940335</td>
<td id="T_53d0e_row6_col16" class="data row6 col16" >349.569763</td>
<td id="T_53d0e_row6_col17" class="data row6 col17" >6144.000000</td>
<td id="T_53d0e_row6_col18" class="data row6 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_6/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row7" class="row_heading level0 row7" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_7/metrics.json</th>
<td id="T_53d0e_row7_col0" class="data row7 col0" >5</td>
<td id="T_53d0e_row7_col1" class="data row7 col1" >128</td>
<td id="T_53d0e_row7_col2" class="data row7 col2" >7</td>
<td id="T_53d0e_row7_col3" class="data row7 col3" >0.073590</td>
<td id="T_53d0e_row7_col4" class="data row7 col4" >9.511523</td>
<td id="T_53d0e_row7_col5" class="data row7 col5" >3.668211</td>
<td id="T_53d0e_row7_col6" class="data row7 col6" >3.599065</td>
<td id="T_53d0e_row7_col7" class="data row7 col7" >13.054041</td>
<td id="T_53d0e_row7_col8" class="data row7 col8" >0.992263</td>
<td id="T_53d0e_row7_col9" class="data row7 col9" >0.992687</td>
<td id="T_53d0e_row7_col10" class="data row7 col10" >140.905991</td>
<td id="T_53d0e_row7_col11" class="data row7 col11" >138.128754</td>
<td id="T_53d0e_row7_col12" class="data row7 col12" >0.970831</td>
<td id="T_53d0e_row7_col13" class="data row7 col13" >31.997070</td>
<td id="T_53d0e_row7_col14" class="data row7 col14" >55.752361</td>
<td id="T_53d0e_row7_col15" class="data row7 col15" >0.928668</td>
<td id="T_53d0e_row7_col16" class="data row7 col16" >499.765717</td>
<td id="T_53d0e_row7_col17" class="data row7 col17" >6144.000000</td>
<td id="T_53d0e_row7_col18" class="data row7 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_7/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row8" class="row_heading level0 row8" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_8/metrics.json</th>
<td id="T_53d0e_row8_col0" class="data row8 col0" >5</td>
<td id="T_53d0e_row8_col1" class="data row8 col1" >128</td>
<td id="T_53d0e_row8_col2" class="data row8 col2" >8</td>
<td id="T_53d0e_row8_col3" class="data row8 col3" >0.089368</td>
<td id="T_53d0e_row8_col4" class="data row8 col4" >7.897106</td>
<td id="T_53d0e_row8_col5" class="data row8 col5" >3.679746</td>
<td id="T_53d0e_row8_col6" class="data row8 col6" >3.599065</td>
<td id="T_53d0e_row8_col7" class="data row8 col7" >11.460873</td>
<td id="T_53d0e_row8_col8" class="data row8 col8" >0.988683</td>
<td id="T_53d0e_row8_col9" class="data row8 col9" >0.989738</td>
<td id="T_53d0e_row8_col10" class="data row8 col10" >157.343246</td>
<td id="T_53d0e_row8_col11" class="data row8 col11" >153.839539</td>
<td id="T_53d0e_row8_col12" class="data row8 col12" >0.968352</td>
<td id="T_53d0e_row8_col13" class="data row8 col13" >31.998373</td>
<td id="T_53d0e_row8_col14" class="data row8 col14" >53.455093</td>
<td id="T_53d0e_row8_col15" class="data row8 col15" >0.913877</td>
<td id="T_53d0e_row8_col16" class="data row8 col16" >732.317871</td>
<td id="T_53d0e_row8_col17" class="data row8 col17" >6144.000000</td>
<td id="T_53d0e_row8_col18" class="data row8 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_8/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row9" class="row_heading level0 row9" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_9/metrics.json</th>
<td id="T_53d0e_row9_col0" class="data row9 col0" >5</td>
<td id="T_53d0e_row9_col1" class="data row9 col1" >128</td>
<td id="T_53d0e_row9_col2" class="data row9 col2" >9</td>
<td id="T_53d0e_row9_col3" class="data row9 col3" >0.096923</td>
<td id="T_53d0e_row9_col4" class="data row9 col4" >5.396312</td>
<td id="T_53d0e_row9_col5" class="data row9 col5" >3.695744</td>
<td id="T_53d0e_row9_col6" class="data row9 col6" >3.599065</td>
<td id="T_53d0e_row9_col7" class="data row9 col7" >8.970472</td>
<td id="T_53d0e_row9_col8" class="data row9 col8" >0.982039</td>
<td id="T_53d0e_row9_col9" class="data row9 col9" >0.982001</td>
<td id="T_53d0e_row9_col10" class="data row9 col10" >181.313721</td>
<td id="T_53d0e_row9_col11" class="data row9 col11" >176.829346</td>
<td id="T_53d0e_row9_col12" class="data row9 col12" >0.966578</td>
<td id="T_53d0e_row9_col13" class="data row9 col13" >31.997070</td>
<td id="T_53d0e_row9_col14" class="data row9 col14" >51.393932</td>
<td id="T_53d0e_row9_col15" class="data row9 col15" >0.895696</td>
<td id="T_53d0e_row9_col16" class="data row9 col16" >1123.425049</td>
<td id="T_53d0e_row9_col17" class="data row9 col17" >6144.000000</td>
<td id="T_53d0e_row9_col18" class="data row9 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_9/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row10" class="row_heading level0 row10" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_10/metrics.json</th>
<td id="T_53d0e_row10_col0" class="data row10 col0" >5</td>
<td id="T_53d0e_row10_col1" class="data row10 col1" >128</td>
<td id="T_53d0e_row10_col2" class="data row10 col2" >10</td>
<td id="T_53d0e_row10_col3" class="data row10 col3" >0.106645</td>
<td id="T_53d0e_row10_col4" class="data row10 col4" >6.193092</td>
<td id="T_53d0e_row10_col5" class="data row10 col5" >3.692219</td>
<td id="T_53d0e_row10_col6" class="data row10 col6" >3.599065</td>
<td id="T_53d0e_row10_col7" class="data row10 col7" >9.754217</td>
<td id="T_53d0e_row10_col8" class="data row10 col8" >0.982780</td>
<td id="T_53d0e_row10_col9" class="data row10 col9" >0.984866</td>
<td id="T_53d0e_row10_col10" class="data row10 col10" >224.287598</td>
<td id="T_53d0e_row10_col11" class="data row10 col11" >218.769226</td>
<td id="T_53d0e_row10_col12" class="data row10 col12" >0.968813</td>
<td id="T_53d0e_row10_col13" class="data row10 col13" >31.998047</td>
<td id="T_53d0e_row10_col14" class="data row10 col14" >47.411495</td>
<td id="T_53d0e_row10_col15" class="data row10 col15" >0.877616</td>
<td id="T_53d0e_row10_col16" class="data row10 col16" >1806.194092</td>
<td id="T_53d0e_row10_col17" class="data row10 col17" >6144.000000</td>
<td id="T_53d0e_row10_col18" class="data row10 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_10/metrics.json</td>
</tr>
<tr>
<th id="T_53d0e_level0_row11" class="row_heading level0 row11" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_11/metrics.json</th>
<td id="T_53d0e_row11_col0" class="data row11 col0" >5</td>
<td id="T_53d0e_row11_col1" class="data row11 col1" >128</td>
<td id="T_53d0e_row11_col2" class="data row11 col2" >11</td>
<td id="T_53d0e_row11_col3" class="data row11 col3" >0.136668</td>
<td id="T_53d0e_row11_col4" class="data row11 col4" >13.087515</td>
<td id="T_53d0e_row11_col5" class="data row11 col5" >3.732962</td>
<td id="T_53d0e_row11_col6" class="data row11 col6" >3.599065</td>
<td id="T_53d0e_row11_col7" class="data row11 col7" >16.484846</td>
<td id="T_53d0e_row11_col8" class="data row11 col8" >0.989557</td>
<td id="T_53d0e_row11_col9" class="data row11 col9" >0.989609</td>
<td id="T_53d0e_row11_col10" class="data row11 col10" >395.539520</td>
<td id="T_53d0e_row11_col11" class="data row11 col11" >391.472504</td>
<td id="T_53d0e_row11_col12" class="data row11 col12" >0.989259</td>
<td id="T_53d0e_row11_col13" class="data row11 col13" >32.000000</td>
<td id="T_53d0e_row11_col14" class="data row11 col14" >31.824055</td>
<td id="T_53d0e_row11_col15" class="data row11 col15" >0.870424</td>
<td id="T_53d0e_row11_col16" class="data row11 col16" >3098.075928</td>
<td id="T_53d0e_row11_col17" class="data row11 col17" >6144.000000</td>
<td id="T_53d0e_row11_col18" class="data row11 col18" >OAI_GPT2Small_v5_128k_resid_post_attn/v5_128k_layer_11/metrics.json</td>
</tr>
</tbody>
</table>