| <?xml version="1.0" encoding="utf-8" standalone="no"?> |
| <!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN" |
| "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd"> |
| <svg xmlns:xlink="http://www.w3.org/1999/xlink" width="964.481pt" height="464.276pt" viewBox="0 0 964.481 464.276" xmlns="http://www.w3.org/2000/svg" version="1.1"> |
| <metadata> |
| <rdf:RDF xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:cc="http://creativecommons.org/ns#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"> |
| <cc:Work> |
| <dc:type rdf:resource="http://purl.org/dc/dcmitype/StillImage"/> |
| <dc:date>2026-10-31T07:43:21.592377</dc:date> |
| <dc:format>image/svg+xml</dc:format> |
| <dc:creator> |
| <cc:Agent> |
| <dc:title>Matplotlib v3.11.1, https://matplotlib.org/</dc:title> |
| </cc:Agent> |
| </dc:creator> |
| </cc:Work> |
| </rdf:RDF> |
| </metadata> |
| <defs> |
| <style type="text/css">*{stroke-linejoin: round; stroke-linecap: butt}</style> |
| </defs> |
| <g id="figure_1"> |
| <g id="patch_1"> |
| <path d="M 0 464.276 |
| L 964.481 464.276 |
| L 964.481 0 |
| L 0 0 |
| z |
| " style="fill: #ffffff"/> |
| </g> |
| <g id="axes_1"> |
| <g id="patch_2"> |
| <path d="M 53.105 310.248 |
| L 539.057838 310.248 |
| L 539.057838 111.528 |
| L 53.105 111.528 |
| z |
| " style="fill: #ffffff"/> |
| </g> |
| <g id="matplotlib.axis_1"> |
| <g id="xtick_1"> |
| <g id="line2d_1"> |
| <defs> |
| <path id="m82583e806c" d="M 0 0 |
| L 0 3.5 |
| " style="stroke: #000000; stroke-width: 0.8"/> |
| </defs> |
| <g> |
| <use xlink:href="#m82583e806c" x="102.997806" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_1"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif" transform="translate(85.367037 343.183346) rotate(-25)">IFEval*</text> |
| </g> |
| </g> |
| <g id="xtick_2"> |
| <g id="line2d_2"> |
| <g> |
| <use xlink:href="#m82583e806c" x="180.231251" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_2"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif" transform="translate(162.165257 343.58821) rotate(-25)">GSM8K</text> |
| </g> |
| </g> |
| <g id="xtick_3"> |
| <g id="line2d_3"> |
| <g> |
| <use xlink:href="#m82583e806c" x="257.464697" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_3"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif" transform="translate(232.071204 350.421947) rotate(-25)">MMLU-Pro</text> |
| </g> |
| </g> |
| <g id="xtick_4"> |
| <g id="line2d_4"> |
| <g> |
| <use xlink:href="#m82583e806c" x="334.698142" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_4"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif" transform="translate(304.906989 354.524307) rotate(-25)">HumanEval</text> |
| </g> |
| </g> |
| <g id="xtick_5"> |
| <g id="line2d_5"> |
| <g> |
| <use xlink:href="#m82583e806c" x="411.931587" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_5"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif" transform="translate(394.415524 343.076371) rotate(-25)">Tools**</text> |
| </g> |
| </g> |
| <g id="xtick_6"> |
| <g id="line2d_6"> |
| <g> |
| <use xlink:href="#m82583e806c" x="489.165033" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_6"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif" transform="translate(467.763259 346.699209) rotate(-25)">TextVQA</text> |
| </g> |
| </g> |
| </g> |
| <g id="matplotlib.axis_2"> |
| <g id="ytick_1"> |
| <g id="line2d_7"> |
| <path d="M 53.105 310.248 |
| L 539.057838 310.248 |
| " clip-path="url(#p773041cdba)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_8"> |
| <defs> |
| <path id="m68bfc73ec2" d="M 0 0 |
| L -3.5 0 |
| " style="stroke: #000000; stroke-width: 0.8"/> |
| </defs> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="53.105" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_7"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="46.105" y="314.806594" transform="rotate(-0 46.105 314.806594)">0</text> |
| </g> |
| </g> |
| <g id="ytick_2"> |
| <g id="line2d_9"> |
| <path d="M 53.105 272.396571 |
| L 539.057838 272.396571 |
| " clip-path="url(#p773041cdba)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_10"> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="53.105" y="272.396571" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_8"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="46.105" y="276.955165" transform="rotate(-0 46.105 276.955165)">20</text> |
| </g> |
| </g> |
| <g id="ytick_3"> |
| <g id="line2d_11"> |
| <path d="M 53.105 234.545143 |
| L 539.057838 234.545143 |
| " clip-path="url(#p773041cdba)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_12"> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="53.105" y="234.545143" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_9"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="46.105" y="239.103737" transform="rotate(-0 46.105 239.103737)">40</text> |
| </g> |
| </g> |
| <g id="ytick_4"> |
| <g id="line2d_13"> |
| <path d="M 53.105 196.693714 |
| L 539.057838 196.693714 |
| " clip-path="url(#p773041cdba)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_14"> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="53.105" y="196.693714" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_10"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="46.105" y="201.252308" transform="rotate(-0 46.105 201.252308)">60</text> |
| </g> |
| </g> |
| <g id="ytick_5"> |
| <g id="line2d_15"> |
| <path d="M 53.105 158.842286 |
| L 539.057838 158.842286 |
| " clip-path="url(#p773041cdba)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_16"> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="53.105" y="158.842286" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_11"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="46.105" y="163.400879" transform="rotate(-0 46.105 163.400879)">80</text> |
| </g> |
| </g> |
| <g id="ytick_6"> |
| <g id="line2d_17"> |
| <path d="M 53.105 120.990857 |
| L 539.057838 120.990857 |
| " clip-path="url(#p773041cdba)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_18"> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="53.105" y="120.990857" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_12"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="46.105" y="125.549451" transform="rotate(-0 46.105 125.549451)">100</text> |
| </g> |
| </g> |
| <g id="text_13"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="16.317188" y="210.888" transform="rotate(-90 16.317188 210.888)">Accuracy (%)</text> |
| </g> |
| </g> |
| <g id="patch_3"> |
| <path d="M 75.193765 310.248 |
| L 93.729792 310.248 |
| L 93.729792 165.150857 |
| L 75.193765 165.150857 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #767676; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_4"> |
| <path d="M 152.427211 310.248 |
| L 170.963238 310.248 |
| L 170.963238 130.453714 |
| L 152.427211 130.453714 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #767676; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_5"> |
| <path d="M 229.660656 310.248 |
| L 248.196683 310.248 |
| L 248.196683 177.768 |
| L 229.660656 177.768 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #767676; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_6"> |
| <path d="M 306.894102 310.248 |
| L 325.430128 310.248 |
| L 325.430128 154.110857 |
| L 306.894102 154.110857 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #767676; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_7"> |
| <path d="M 384.127547 310.248 |
| L 402.663574 310.248 |
| L 402.663574 139.916571 |
| L 384.127547 139.916571 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #767676; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_8"> |
| <path d="M 461.360992 310.248 |
| L 479.897019 310.248 |
| L 479.897019 139.916571 |
| L 461.360992 139.916571 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #767676; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_9"> |
| <path d="M 93.729792 310.248 |
| L 112.265819 310.248 |
| L 112.265819 168.305143 |
| L 93.729792 168.305143 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #3775ba; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_10"> |
| <path d="M 170.963238 310.248 |
| L 189.499265 310.248 |
| L 189.499265 125.722286 |
| L 170.963238 125.722286 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #3775ba; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_11"> |
| <path d="M 248.196683 310.248 |
| L 266.73271 310.248 |
| L 266.73271 180.922286 |
| L 248.196683 180.922286 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #3775ba; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_12"> |
| <path d="M 325.430128 310.248 |
| L 343.966155 310.248 |
| L 343.966155 158.842286 |
| L 325.430128 158.842286 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #3775ba; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_13"> |
| <path d="M 402.663574 310.248 |
| L 421.199601 310.248 |
| L 421.199601 139.916571 |
| L 402.663574 139.916571 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #3775ba; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_14"> |
| <path d="M 479.897019 310.248 |
| L 498.433046 310.248 |
| L 498.433046 144.648 |
| L 479.897019 144.648 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #3775ba; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_15"> |
| <path d="M 112.265819 310.248 |
| L 130.801846 310.248 |
| L 130.801846 155.688 |
| L 112.265819 155.688 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #0f4d92; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_16"> |
| <path d="M 189.499265 310.248 |
| L 208.035291 310.248 |
| L 208.035291 128.088 |
| L 189.499265 128.088 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #0f4d92; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_17"> |
| <path d="M 266.73271 310.248 |
| L 285.268737 310.248 |
| L 285.268737 199.848 |
| L 266.73271 199.848 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #0f4d92; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_18"> |
| <path d="M 343.966155 310.248 |
| L 362.502182 310.248 |
| L 362.502182 163.573714 |
| L 343.966155 163.573714 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #0f4d92; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_19"> |
| <path d="M 421.199601 310.248 |
| L 439.735628 310.248 |
| L 439.735628 139.916571 |
| L 421.199601 139.916571 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #0f4d92; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_20"> |
| <path d="M 498.433046 310.248 |
| L 516.969073 310.248 |
| L 516.969073 135.185143 |
| L 498.433046 135.185143 |
| z |
| " clip-path="url(#p773041cdba)" style="fill: #0f4d92; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="patch_21"> |
| <path d="M 53.105 310.248 |
| L 53.105 111.528 |
| " style="fill: none; stroke: #000000; stroke-width: 1.8; stroke-linejoin: miter; stroke-linecap: square"/> |
| </g> |
| <g id="patch_22"> |
| <path d="M 53.105 310.248 |
| L 539.057838 310.248 |
| " style="fill: none; stroke: #000000; stroke-width: 1.8; stroke-linejoin: miter; stroke-linecap: square"/> |
| </g> |
| <g id="text_14"> |
| <text style="font-size: 14px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="296.081419" y="105.528" transform="rotate(-0 296.081419 105.528)">A First complete regression run</text> |
| </g> |
| </g> |
| <g id="axes_2"> |
| <g id="patch_23"> |
| <path d="M 653.560476 310.248 |
| L 957.281 310.248 |
| L 957.281 111.528 |
| L 653.560476 111.528 |
| z |
| " style="fill: #ffffff"/> |
| </g> |
| <g id="matplotlib.axis_3"> |
| <g id="xtick_7"> |
| <g id="line2d_19"> |
| <path d="M 653.560476 310.248 |
| L 653.560476 111.528 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_20"> |
| <g> |
| <use xlink:href="#m82583e806c" x="653.560476" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_15"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="653.560476" y="326.365187" transform="rotate(-0 653.560476 326.365187)">−6</text> |
| </g> |
| </g> |
| <g id="xtick_8"> |
| <g id="line2d_21"> |
| <path d="M 721.053926 310.248 |
| L 721.053926 111.528 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_22"> |
| <g> |
| <use xlink:href="#m82583e806c" x="721.053926" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_16"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="721.053926" y="326.365187" transform="rotate(-0 721.053926 326.365187)">−4</text> |
| </g> |
| </g> |
| <g id="xtick_9"> |
| <g id="line2d_23"> |
| <path d="M 788.547376 310.248 |
| L 788.547376 111.528 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_24"> |
| <g> |
| <use xlink:href="#m82583e806c" x="788.547376" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_17"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="788.547376" y="326.365187" transform="rotate(-0 788.547376 326.365187)">−2</text> |
| </g> |
| </g> |
| <g id="xtick_10"> |
| <g id="line2d_25"> |
| <path d="M 856.040825 310.248 |
| L 856.040825 111.528 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_26"> |
| <g> |
| <use xlink:href="#m82583e806c" x="856.040825" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_18"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="856.040825" y="326.365187" transform="rotate(-0 856.040825 326.365187)">0</text> |
| </g> |
| </g> |
| <g id="xtick_11"> |
| <g id="line2d_27"> |
| <path d="M 923.534275 310.248 |
| L 923.534275 111.528 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke: #b0b0b0; stroke-opacity: 0.16; stroke-width: 0.8; stroke-linecap: square"/> |
| </g> |
| <g id="line2d_28"> |
| <g> |
| <use xlink:href="#m82583e806c" x="923.534275" y="310.248" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_19"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="923.534275" y="326.365187" transform="rotate(-0 923.534275 326.365187)">2</text> |
| </g> |
| </g> |
| <g id="text_20"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="805.420738" y="342.366125" transform="rotate(-0 805.420738 342.366125)">Percentage points (paired-bootstrap 95% CI)</text> |
| </g> |
| </g> |
| <g id="matplotlib.axis_4"> |
| <g id="ytick_7"> |
| <g id="line2d_29"> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="653.560476" y="163.573714" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_21"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="646.560476" y="168.132308" transform="rotate(-0 646.560476 168.132308)">v2 exact</text> |
| </g> |
| </g> |
| <g id="ytick_8"> |
| <g id="line2d_30"> |
| <g> |
| <use xlink:href="#m68bfc73ec2" x="653.560476" y="258.202286" style="stroke: #000000; stroke-width: 0.8"/> |
| </g> |
| </g> |
| <g id="text_22"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: end" x="646.560476" y="262.761348" transform="rotate(-0 646.560476 262.761348)">v2 fast</text> |
| </g> |
| </g> |
| </g> |
| <g id="LineCollection_1"> |
| <path d="M 754.800651 163.573714 |
| L 912.285367 163.573714 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke: #3775ba; stroke-width: 2"/> |
| </g> |
| <g id="line2d_31"> |
| <defs> |
| <path id="m2d91ce3a39" d="M 0 5 |
| L 0 -5 |
| " style="stroke: #3775ba"/> |
| </defs> |
| <g clip-path="url(#pef5edb4016)"> |
| <use xlink:href="#m2d91ce3a39" x="754.800651" y="163.573714" style="fill: #3775ba; stroke: #3775ba"/> |
| </g> |
| </g> |
| <g id="line2d_32"> |
| <g clip-path="url(#pef5edb4016)"> |
| <use xlink:href="#m2d91ce3a39" x="912.285367" y="163.573714" style="fill: #3775ba; stroke: #3775ba"/> |
| </g> |
| </g> |
| <g id="LineCollection_2"> |
| <path d="M 709.805017 258.202286 |
| L 912.285367 258.202286 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke: #0f4d92; stroke-width: 2"/> |
| </g> |
| <g id="line2d_33"> |
| <defs> |
| <path id="mdcc8c3e22a" d="M 0 5 |
| L 0 -5 |
| " style="stroke: #0f4d92"/> |
| </defs> |
| <g clip-path="url(#pef5edb4016)"> |
| <use xlink:href="#mdcc8c3e22a" x="709.805017" y="258.202286" style="fill: #0f4d92; stroke: #0f4d92"/> |
| </g> |
| </g> |
| <g id="line2d_34"> |
| <g clip-path="url(#pef5edb4016)"> |
| <use xlink:href="#mdcc8c3e22a" x="912.285367" y="258.202286" style="fill: #0f4d92; stroke: #0f4d92"/> |
| </g> |
| </g> |
| <g id="line2d_35"> |
| <path d="M 856.040825 310.248 |
| L 856.040825 111.528 |
| " clip-path="url(#pef5edb4016)" style="fill: none; stroke-dasharray: 3.7,1.6; stroke-dashoffset: 0; stroke: #767676"/> |
| </g> |
| <g id="line2d_36"> |
| <defs> |
| <path id="m966702bc1a" d="M 0 3 |
| C 0.795609 3 1.55874 2.683901 2.12132 2.12132 |
| C 2.683901 1.55874 3 0.795609 3 0 |
| C 3 -0.795609 2.683901 -1.55874 2.12132 -2.12132 |
| C 1.55874 -2.683901 0.795609 -3 0 -3 |
| C -0.795609 -3 -1.55874 -2.683901 -2.12132 -2.12132 |
| C -2.683901 -1.55874 -3 -0.795609 -3 0 |
| C -3 0.795609 -2.683901 1.55874 -2.12132 2.12132 |
| C -1.55874 2.683901 -0.795609 3 0 3 |
| z |
| " style="stroke: #3775ba"/> |
| </defs> |
| <g clip-path="url(#pef5edb4016)"> |
| <use xlink:href="#m966702bc1a" x="833.543009" y="163.573714" style="fill: #3775ba; stroke: #3775ba"/> |
| </g> |
| </g> |
| <g id="line2d_37"> |
| <defs> |
| <path id="md02f69db95" d="M 0 3 |
| C 0.795609 3 1.55874 2.683901 2.12132 2.12132 |
| C 2.683901 1.55874 3 0.795609 3 0 |
| C 3 -0.795609 2.683901 -1.55874 2.12132 -2.12132 |
| C 1.55874 -2.683901 0.795609 -3 0 -3 |
| C -0.795609 -3 -1.55874 -2.683901 -2.12132 -2.12132 |
| C -2.683901 -1.55874 -3 -0.795609 -3 0 |
| C -3 0.795609 -2.683901 1.55874 -2.12132 2.12132 |
| C -1.55874 2.683901 -0.795609 3 0 3 |
| z |
| " style="stroke: #0f4d92"/> |
| </defs> |
| <g clip-path="url(#pef5edb4016)"> |
| <use xlink:href="#md02f69db95" x="811.045192" y="258.202286" style="fill: #0f4d92; stroke: #0f4d92"/> |
| </g> |
| </g> |
| <g id="patch_24"> |
| <path d="M 653.560476 310.248 |
| L 653.560476 111.528 |
| " style="fill: none; stroke: #000000; stroke-width: 1.8; stroke-linejoin: miter; stroke-linecap: square"/> |
| </g> |
| <g id="patch_25"> |
| <path d="M 653.560476 310.248 |
| L 957.281 310.248 |
| " style="fill: none; stroke: #000000; stroke-width: 1.8; stroke-linejoin: miter; stroke-linecap: square"/> |
| </g> |
| <g id="text_23"> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: start" x="670.433838" y="184.392" transform="rotate(-0 670.433838 184.392)">-0.67 pp [-3.00, +1.67]</text> |
| </g> |
| <g id="text_24"> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: start" x="670.433838" y="279.020571" transform="rotate(-0 670.433838 279.020571)">-1.33 pp [-4.33, +1.67]</text> |
| </g> |
| <g id="text_25"> |
| <text style="font-size: 14px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="805.420738" y="105.528" transform="rotate(-0 805.420738 105.528)">B Pooled change vs frozen v1</text> |
| </g> |
| </g> |
| <g id="text_26"> |
| <text style="font-weight: 700; font-size: 18px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle" x="480.353" y="20.877187" transform="rotate(-0 480.353 20.877187)">Runtime v2 quality — preserved evidence, explicit trade-offs</text> |
| </g> |
| <g id="text_27"> |
| <text style="font-size: 12px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: middle; fill: #4d4d4d" x="480.353" y="39.9888" transform="rotate(-0 480.353 39.9888)">Same 300-item suite • thinking off • greedy • evaluation concurrency 16</text> |
| </g> |
| <g id="text_28"> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; fill: #4d4d4d" transform="translate(292.342844 383.715844)">Frozen: 251/300 (83.7%) • exact: 249/300 (83.0%) • fast: 247/300 (82.3%)</text> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; fill: #4d4d4d" transform="translate(262.981125 400.335375)">Fast MMLU-Pro first run: 58.3% vs 70.0% frozen (−11.7 pp); fast reruns: 71.7%, 65.0%.</text> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; fill: #4d4d4d" transform="translate(225.64675 417.335375)">Fast truncations: 20 vs 14 frozen. Fast fails original first-run per-task/stability tripwires; not “lossless.”</text> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; fill: #4d4d4d" transform="translate(215.843625 434.335375)">*Re-implemented IFEval checker. **20 hand-written tool cases. Small subsets, not full-benchmark results.</text> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; fill: #4d4d4d" transform="translate(237.64675 451.196312)">Source: runtime-v2/runs/final_*/eval; all reruns retained in runtime-v2/runs/var and RESULTS.md.</text> |
| </g> |
| <g id="legend_1"> |
| <g id="patch_26"> |
| <path d="M 302.414719 80.930437 |
| L 322.414719 80.930437 |
| L 322.414719 73.930437 |
| L 302.414719 73.930437 |
| z |
| " style="fill: #767676; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="text_29"> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: start" x="330.414719" y="80.930437" transform="rotate(-0 330.414719 80.930437)">Frozen v1 + MTP</text> |
| </g> |
| <g id="patch_27"> |
| <path d="M 434.147531 80.930437 |
| L 454.147531 80.930437 |
| L 454.147531 73.930437 |
| L 434.147531 73.930437 |
| z |
| " style="fill: #3775ba; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="text_30"> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: start" x="462.147531" y="80.930437" transform="rotate(-0 462.147531 80.930437)">v2 exact + MTP</text> |
| </g> |
| <g id="patch_28"> |
| <path d="M 560.552219 80.930437 |
| L 580.552219 80.930437 |
| L 580.552219 73.930437 |
| L 560.552219 73.930437 |
| z |
| " style="fill: #0f4d92; stroke: #000000; stroke-width: 0.7; stroke-linejoin: miter"/> |
| </g> |
| <g id="text_31"> |
| <text style="font-size: 10px; font-family: 'DejaVu Sans', 'Arial', 'Helvetica', sans-serif; text-anchor: start" x="588.552219" y="80.930437" transform="rotate(-0 588.552219 80.930437)">v2 fast + MTP</text> |
| </g> |
| </g> |
| </g> |
| <defs> |
| <clipPath id="p773041cdba"> |
| <rect x="53.105" y="111.528" width="485.952838" height="198.72"/> |
| </clipPath> |
| <clipPath id="pef5edb4016"> |
| <rect x="653.560476" y="111.528" width="303.720524" height="198.72"/> |
| </clipPath> |
| </defs> |
| </svg> |
|
|
