github/microsoft-SkillOpt
1
0
Fork 0
mirror of https://github.com/microsoft/SkillOpt.git synced 2026-07-03 14:02:58 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add main results method comparison chart

Browse Source
This commit is contained in:
Lliar-liar
2026-05-24 17:55:22 +00:00
parent c64fbcd4f8
commit 9012a79827
3 changed files with 44 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
index.html
View File

@@ -458,6 +458,21 @@
background: #ffffff;
}
.comparison-frame {
margin-top: 18px;
background: #0b1018;
border: 1px solid var(--line-strong);
border-radius: 8px;
overflow: hidden;
box-shadow: var(--shadow);
}
.comparison-frame img {
display: block;
width: 100%;
background: #0b1018;
}
.caption {
padding: 13px 16px;
color: var(--muted);
@@ -1459,6 +1474,13 @@
</table>
</div>
<figure class="comparison-frame">
<img src="skillopt-assets/main-results-comparison.png" alt="Bar charts comparing SkillOpt with no skill, human skill, LLM skill, Trace2Skill, TextGrad, and GEPA across SearchQA, SpreadsheetBench, OfficeQA, DocVQA, LiveMath, and ALFWorld.">
<figcaption class="caption">
Method comparison from the project video. Bars report per-benchmark direct-chat accuracy averaged over seven target models; SkillOpt is best or tied-best in every panel.
</figcaption>
</figure>
</section>
<section class="section" id="ablations">

BIN
skillopt-assets/main-results-comparison.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 644 KiB

22
skillopt.html
View File

@@ -458,6 +458,21 @@
background: #ffffff;
}
.comparison-frame {
margin-top: 18px;
background: #0b1018;
border: 1px solid var(--line-strong);
border-radius: 8px;
overflow: hidden;
box-shadow: var(--shadow);
}
.comparison-frame img {
display: block;
width: 100%;
background: #0b1018;
}
.caption {
padding: 13px 16px;
color: var(--muted);
@@ -1459,6 +1474,13 @@
</table>
</div>
<figure class="comparison-frame">
<img src="skillopt-assets/main-results-comparison.png" alt="Bar charts comparing SkillOpt with no skill, human skill, LLM skill, Trace2Skill, TextGrad, and GEPA across SearchQA, SpreadsheetBench, OfficeQA, DocVQA, LiveMath, and ALFWorld.">
<figcaption class="caption">
Method comparison from the project video. Bars report per-benchmark direct-chat accuracy averaged over seven target models; SkillOpt is best or tied-best in every panel.
</figcaption>
</figure>
</section>
<section class="section" id="ablations">