@@ -806,38 +806,38 @@ <h2 class="title is-3 has-text-centered">Evaluation Framework</h2>
806806 </ div >
807807 < div class ="content has-text-justified ">
808808 < p > Our evaluation framework employs four core metrics:</ p >
809- < div class ="columns " style ="margin-top: 10px ">
809+ < div class ="columns " style ="margin-top: 10px; align-items: stretch; ">
810810 < div class ="column is-3 ">
811- < div class ="box ">
811+ < div class ="box " style =" height: 100%; display: flex; flex-direction: column; " >
812812 < h5 class ="title is-5 "> 🔧 Render Score (T/V)</ h5 >
813- < p style ="text-align: left ">
813+ < p style ="text-align: left; flex: 1; ">
814814 Binary metric (0 or 1) indicating whether the generated code can be
815815 successfully loaded or rendered without syntax errors
816816 </ p >
817817 </ div >
818818 </ div >
819819 < div class ="column is-3 ">
820- < div class ="box ">
820+ < div class ="box " style =" height: 100%; display: flex; flex-direction: column; " >
821821 < h5 class ="title is-5 "> ✓ Syntax Score (T)</ h5 >
822- < p style ="text-align: left ">
822+ < p style ="text-align: left; flex: 1; ">
823823 Verifies structural correctness using format-specific
824824 parsers or rendering engines
825825 </ p >
826826 </ div >
827827 </ div >
828828 < div class ="column is-3 ">
829- < div class ="box ">
829+ < div class ="box " style =" height: 100%; display: flex; flex-direction: column; " >
830830 < h5 class ="title is-5 "> 🔍 Keyword Matching (V)</ h5 >
831- < p style ="text-align: left ">
831+ < p style ="text-align: left; flex: 1; ">
832832 Evaluates presence of required structural elements using
833833 path-based rules
834834 </ p >
835835 </ div >
836836 </ div >
837837 < div class ="column is-3 ">
838- < div class ="box ">
838+ < div class ="box " style =" height: 100%; display: flex; flex-direction: column; " >
839839 < h5 class ="title is-5 "> 👁️ VQA Score (V)</ h5 >
840- < p style ="text-align: left ">
840+ < p style ="text-align: left; flex: 1; ">
841841 Assesses visual correctness through GPT-4V-based question
842842 answering (V tasks only)
843843 </ p >
@@ -847,29 +847,29 @@ <h5 class="title is-5">👁️ VQA Score (V)</h5>
847847
848848 < div style ="margin-top: 40px ">
849849 < h3 class ="title is-4 has-text-centered "> Score Aggregation Methods</ h3 >
850- < div class ="columns " style ="margin-top: 20px ">
850+ < div class ="columns " style ="margin-top: 20px; align-items: stretch; ">
851851 < div class ="column is-6 ">
852- < div class ="box has-background-info-light ">
852+ < div class ="box has-background-info-light " style =" height: 100%; display: flex; flex-direction: column; " >
853853 < h5 class ="title is-5 has-text-centered "> 🎨 Renderable Formats</ h5 >
854854 < p class ="has-text-centered " style ="margin-bottom: 15px ">
855855 < em > HTML, React, SVG, LaTeX, Mermaid, etc.</ em >
856856 </ p >
857- < div class ="content ">
858- < pre style ="background: #f5f5f5; padding: 15px; border-radius: 5px; font-size: 0.9em; "> final_score = (0.2 × render_score ) +
859- (0.1 × keyword_matching ) +
860- (0.7 × vqa_score )</ pre >
857+ < div class ="content " style =" flex: 1; " >
858+ < pre style ="background: #f5f5f5; padding: 15px; border-radius: 5px; font-size: 0.9em; "> Final Score = (0.2 × Render Score ) +
859+ (0.1 × Keyword Matching ) +
860+ (0.7 × VQA Score )</ pre >
861861 </ div >
862862 </ div >
863863 </ div >
864864 < div class ="column is-6 ">
865- < div class ="box has-background-light ">
865+ < div class ="box has-background-light " style =" height: 100%; display: flex; flex-direction: column; " >
866866 < h5 class ="title is-5 has-text-centered "> 📝 Non-Renderable Formats</ h5 >
867867 < p class ="has-text-centered " style ="margin-bottom: 15px ">
868868 < em > JSON, XML, YAML, CSV, TOML</ em >
869869 </ p >
870- < div class ="content ">
871- < pre style ="background: #f5f5f5; padding: 15px; border-radius: 5px; font-size: 0.9em; "> final_score = (0.2 × render_score ) +
872- (0.8 × syntax_score )</ pre >
870+ < div class ="content " style =" flex: 1; " >
871+ < pre style ="background: #f5f5f5; padding: 15px; border-radius: 5px; font-size: 0.9em; "> Final Score = (0.2 × Render Score ) +
872+ (0.8 × Syntax Score )</ pre >
873873 </ div >
874874 </ div >
875875 </ div >
0 commit comments