Datasets:
study large_stringclasses 8
values | benchmark large_stringclasses 35
values | model large_stringclasses 30
values | ma_configuration large_stringclasses 26
values | baseline_rung large_stringclasses 3
values | baseline large_stringclasses 12
values | key large_stringlengths 26 100 | accuracy_difference float64 -0.75 0.48 | cost_ratio float64 0.21 225 ⌀ | p_base float64 0.01 0.99 | p_mas float64 0 0.97 | n_items float64 20 14k |
|---|---|---|---|---|---|---|---|---|---|---|---|
Jwalapuram | GPQA-Diamond | GPT-4o | DyLAN | A | CoT | Jwalapuram|GPQA-Diamond|GPT-4o|DyLAN | -0.004 | 5.642857 | 0.5341 | 0.5301 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | DyLAN | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-4o|DyLAN | -0.016 | 0.626984 | 0.5461 | 0.5301 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | MAS-Zero | A | CoT | Jwalapuram|GPQA-Diamond|GPT-4o|MAS-Zero | -0.1004 | 185.571429 | 0.5341 | 0.4337 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | MAS-Zero | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-4o|MAS-Zero | -0.1124 | 20.619048 | 0.5461 | 0.4337 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | AFlow | A | CoT | Jwalapuram|GPQA-Diamond|GPT-4o|AFlow | -0.014 | 93.5 | 0.5341 | 0.5201 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | AFlow | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-4o|AFlow | -0.026 | 10.388889 | 0.5461 | 0.5201 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | MaAS | A | CoT | Jwalapuram|GPQA-Diamond|GPT-4o|MaAS | -0.0201 | 7.785714 | 0.5341 | 0.514 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | MaAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-4o|MaAS | -0.0321 | 0.865079 | 0.5461 | 0.514 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | ADAS | A | CoT | Jwalapuram|GPQA-Diamond|GPT-4o|ADAS | -0.0901 | 4.285714 | 0.5341 | 0.444 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | ADAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-4o|ADAS | -0.1021 | 0.47619 | 0.5461 | 0.444 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | MAS-Orchestra | A | CoT | Jwalapuram|GPQA-Diamond|GPT-4o|MAS-Orchestra | 0.0081 | 6.642857 | 0.5341 | 0.5422 | 166 |
Jwalapuram | GPQA-Diamond | GPT-4o | MAS-Orchestra | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-4o|MAS-Orchestra | -0.0039 | 0.738095 | 0.5461 | 0.5422 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | DyLAN | A | CoT | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|DyLAN | 0.0522 | 4.736842 | 0.7048 | 0.757 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | DyLAN | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|DyLAN | 0.0422 | 0.473684 | 0.7148 | 0.757 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | MAS-Zero | A | CoT | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|MAS-Zero | 0.0422 | 73.263158 | 0.7048 | 0.747 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | MAS-Zero | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|MAS-Zero | 0.0322 | 7.326316 | 0.7148 | 0.747 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | AFlow | A | CoT | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|AFlow | 0.01 | 54.736842 | 0.7048 | 0.7148 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | AFlow | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|AFlow | 0 | 5.473684 | 0.7148 | 0.7148 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | MaAS | A | CoT | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|MaAS | 0.014 | 3.631579 | 0.7048 | 0.7188 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | MaAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|MaAS | 0.004 | 0.363158 | 0.7148 | 0.7188 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | ADAS | A | CoT | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|ADAS | 0.01 | 18.421053 | 0.7048 | 0.7148 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | ADAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|ADAS | 0 | 1.842105 | 0.7148 | 0.7148 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | MAS-Orchestra | A | CoT | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|MAS-Orchestra | 0.014 | 3.684211 | 0.7048 | 0.7188 | 166 |
Jwalapuram | GPQA-Diamond | GPT-OSS-120B | MAS-Orchestra | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-OSS-120B|MAS-Orchestra | 0.004 | 0.368421 | 0.7148 | 0.7188 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | DyLAN | A | CoT | Jwalapuram|GPQA-Diamond|GPT-5|DyLAN | -0.0481 | 5.19337 | 0.8714 | 0.8233 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | DyLAN | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-5|DyLAN | -0.0502 | 0.81052 | 0.8735 | 0.8233 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | MAS-Zero | A | CoT | Jwalapuram|GPQA-Diamond|GPT-5|MAS-Zero | -0.0039 | 87.955801 | 0.8714 | 0.8675 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | MAS-Zero | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-5|MAS-Zero | -0.006 | 13.727096 | 0.8735 | 0.8675 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | AFlow | A | CoT | Jwalapuram|GPQA-Diamond|GPT-5|AFlow | -0.0301 | 37.928177 | 0.8714 | 0.8413 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | AFlow | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-5|AFlow | -0.0322 | 5.919379 | 0.8735 | 0.8413 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | MaAS | A | CoT | Jwalapuram|GPQA-Diamond|GPT-5|MaAS | -0.0019 | 5.31768 | 0.8714 | 0.8695 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | MaAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-5|MaAS | -0.004 | 0.82992 | 0.8735 | 0.8695 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | ADAS | A | CoT | Jwalapuram|GPQA-Diamond|GPT-5|ADAS | -0.0191 | 114.930939 | 0.8714 | 0.8523 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | ADAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-5|ADAS | -0.0212 | 17.937055 | 0.8735 | 0.8523 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | MAS-Orchestra | A | CoT | Jwalapuram|GPQA-Diamond|GPT-5|MAS-Orchestra | -0.016 | 5.759669 | 0.8714 | 0.8554 | 166 |
Jwalapuram | GPQA-Diamond | GPT-5 | MAS-Orchestra | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|GPT-5|MAS-Orchestra | -0.0181 | 0.898901 | 0.8735 | 0.8554 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | DyLAN | A | CoT | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|DyLAN | 0.1024 | 3.176341 | 0.7711 | 0.8735 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | DyLAN | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|DyLAN | 0.0422 | 0.657512 | 0.8313 | 0.8735 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | MAS-Zero | A | CoT | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|MAS-Zero | 0.0903 | 29.076838 | 0.7711 | 0.8614 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | MAS-Zero | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|MAS-Zero | 0.0301 | 6.018995 | 0.8313 | 0.8614 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | AFlow | A | CoT | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|AFlow | 0.0622 | 12.757877 | 0.7711 | 0.8333 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | AFlow | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|AFlow | 0.002 | 2.64092 | 0.8313 | 0.8333 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | MaAS | A | CoT | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|MaAS | 0.0984 | 5.242123 | 0.7711 | 0.8695 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | MaAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|MaAS | 0.0382 | 1.085136 | 0.8313 | 0.8695 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | ADAS | A | CoT | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|ADAS | 0.0741 | 38.363737 | 0.7711 | 0.8452 | 166 |
Jwalapuram | GPQA-Diamond | Gemini-2.5-Pro | ADAS | B | CoT-SC (5) | Jwalapuram|GPQA-Diamond|Gemini-2.5-Pro|ADAS | 0.0139 | 7.941412 | 0.8313 | 0.8452 | 166 |
Jwalapuram | HLE-Maths | GPT-4o | DyLAN | A | CoT | Jwalapuram|HLE-Maths|GPT-4o|DyLAN | -0.0103 | 9.923077 | 0.046 | 0.0357 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | DyLAN | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-4o|DyLAN | -0.0119 | 1.482759 | 0.0476 | 0.0357 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | MAS-Zero | A | CoT | Jwalapuram|HLE-Maths|GPT-4o|MAS-Zero | -0.0143 | 192.153846 | 0.046 | 0.0317 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | MAS-Zero | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-4o|MAS-Zero | -0.0159 | 28.712644 | 0.0476 | 0.0317 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | AFlow | A | CoT | Jwalapuram|HLE-Maths|GPT-4o|AFlow | -0.0183 | 225.076923 | 0.046 | 0.0277 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | AFlow | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-4o|AFlow | -0.0199 | 33.632184 | 0.0476 | 0.0277 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | MaAS | A | CoT | Jwalapuram|HLE-Maths|GPT-4o|MaAS | -0.0023 | 4.607692 | 0.046 | 0.0437 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | MaAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-4o|MaAS | -0.0039 | 0.688506 | 0.0476 | 0.0437 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | ADAS | A | CoT | Jwalapuram|HLE-Maths|GPT-4o|ADAS | -0.008 | 38.923077 | 0.046 | 0.038 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | ADAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-4o|ADAS | -0.0096 | 5.816092 | 0.0476 | 0.038 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | MAS-Orchestra | A | CoT | Jwalapuram|HLE-Maths|GPT-4o|MAS-Orchestra | -0.0123 | 7.153846 | 0.046 | 0.0337 | 168 |
Jwalapuram | HLE-Maths | GPT-4o | MAS-Orchestra | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-4o|MAS-Orchestra | -0.0139 | 1.068966 | 0.0476 | 0.0337 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | DyLAN | A | CoT | Jwalapuram|HLE-Maths|GPT-OSS-120B|DyLAN | 0.0059 | 5.333333 | 0.0873 | 0.0932 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | DyLAN | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-OSS-120B|DyLAN | -0.0457 | 0.410256 | 0.1389 | 0.0932 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | MAS-Zero | A | CoT | Jwalapuram|HLE-Maths|GPT-OSS-120B|MAS-Zero | 0.0702 | 52.066667 | 0.0873 | 0.1575 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | MAS-Zero | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-OSS-120B|MAS-Zero | 0.0186 | 4.005128 | 0.1389 | 0.1575 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | AFlow | A | CoT | Jwalapuram|HLE-Maths|GPT-OSS-120B|AFlow | 0.0078 | 76.666667 | 0.0873 | 0.0951 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | AFlow | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-OSS-120B|AFlow | -0.0438 | 5.897436 | 0.1389 | 0.0951 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | MaAS | A | CoT | Jwalapuram|HLE-Maths|GPT-OSS-120B|MaAS | 0.0675 | 8.666667 | 0.0873 | 0.1548 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | MaAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-OSS-120B|MaAS | 0.0159 | 0.666667 | 0.1389 | 0.1548 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | ADAS | A | CoT | Jwalapuram|HLE-Maths|GPT-OSS-120B|ADAS | -0.008 | 6.333333 | 0.0873 | 0.0793 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | ADAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-OSS-120B|ADAS | -0.0596 | 0.487179 | 0.1389 | 0.0793 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | MAS-Orchestra | A | CoT | Jwalapuram|HLE-Maths|GPT-OSS-120B|MAS-Orchestra | 0.0059 | 6.666667 | 0.0873 | 0.0932 | 168 |
Jwalapuram | HLE-Maths | GPT-OSS-120B | MAS-Orchestra | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-OSS-120B|MAS-Orchestra | -0.0457 | 0.512821 | 0.1389 | 0.0932 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | DyLAN | A | CoT | Jwalapuram|HLE-Maths|GPT-5|DyLAN | 0.0564 | 3.435821 | 0.2987 | 0.3551 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | DyLAN | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-5|DyLAN | 0.0159 | 0.990534 | 0.3392 | 0.3551 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | MAS-Zero | A | CoT | Jwalapuram|HLE-Maths|GPT-5|MAS-Zero | 0.0833 | 38.456716 | 0.2987 | 0.382 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | MAS-Zero | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-5|MAS-Zero | 0.0428 | 11.086919 | 0.3392 | 0.382 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | AFlow | A | CoT | Jwalapuram|HLE-Maths|GPT-5|AFlow | 0.0286 | 23.420896 | 0.2987 | 0.3273 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | AFlow | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-5|AFlow | -0.0119 | 6.752151 | 0.3392 | 0.3273 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | MaAS | A | CoT | Jwalapuram|HLE-Maths|GPT-5|MaAS | 0.0545 | 6.602985 | 0.2987 | 0.3532 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | MaAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-5|MaAS | 0.014 | 1.903614 | 0.3392 | 0.3532 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | ADAS | A | CoT | Jwalapuram|HLE-Maths|GPT-5|ADAS | 0.0473 | 25.280597 | 0.2987 | 0.346 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | ADAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-5|ADAS | 0.0068 | 7.288296 | 0.3392 | 0.346 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | MAS-Orchestra | A | CoT | Jwalapuram|HLE-Maths|GPT-5|MAS-Orchestra | 0.0777 | 4.402687 | 0.2987 | 0.3764 | 168 |
Jwalapuram | HLE-Maths | GPT-5 | MAS-Orchestra | B | CoT-SC (5) | Jwalapuram|HLE-Maths|GPT-5|MAS-Orchestra | 0.0372 | 1.269277 | 0.3392 | 0.3764 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | DyLAN | A | CoT | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|DyLAN | 0.0654 | 8.320687 | 0.256 | 0.3214 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | DyLAN | B | CoT-SC (5) | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|DyLAN | 0.0595 | 1.628235 | 0.2619 | 0.3214 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | MAS-Zero | A | CoT | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|MAS-Zero | 0.039 | null | 0.256 | 0.295 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | MAS-Zero | B | CoT-SC (5) | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|MAS-Zero | 0.0331 | null | 0.2619 | 0.295 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | AFlow | A | CoT | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|AFlow | -0.0655 | 6.342276 | 0.256 | 0.1905 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | AFlow | B | CoT-SC (5) | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|AFlow | -0.0714 | 1.241089 | 0.2619 | 0.1905 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | MaAS | A | CoT | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|MaAS | 0.0892 | 5.21442 | 0.256 | 0.3452 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | MaAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|MaAS | 0.0833 | 1.020385 | 0.2619 | 0.3452 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | ADAS | A | CoT | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|ADAS | 0.0238 | 39.740096 | 0.256 | 0.2798 | 168 |
Jwalapuram | HLE-Maths | Gemini-2.5-Pro | ADAS | B | CoT-SC (5) | Jwalapuram|HLE-Maths|Gemini-2.5-Pro|ADAS | 0.0179 | 7.776547 | 0.2619 | 0.2798 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | DyLAN | A | CoT | Jwalapuram|SWE-Bench Lite|GPT-4o|DyLAN | -0.0074 | 9.015625 | 0.2002 | 0.1928 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | DyLAN | B | CoT-SC (5) | Jwalapuram|SWE-Bench Lite|GPT-4o|DyLAN | -0.0273 | 1.371361 | 0.2201 | 0.1928 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | MAS-Zero | A | CoT | Jwalapuram|SWE-Bench Lite|GPT-4o|MAS-Zero | -0.0485 | 47.300781 | 0.2002 | 0.1517 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | MAS-Zero | B | CoT-SC (5) | Jwalapuram|SWE-Bench Lite|GPT-4o|MAS-Zero | -0.0684 | 7.19489 | 0.2201 | 0.1517 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | AFlow | A | CoT | Jwalapuram|SWE-Bench Lite|GPT-4o|AFlow | -0.0945 | 35.28125 | 0.2002 | 0.1057 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | AFlow | B | CoT-SC (5) | Jwalapuram|SWE-Bench Lite|GPT-4o|AFlow | -0.1144 | 5.366607 | 0.2201 | 0.1057 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | MaAS | A | CoT | Jwalapuram|SWE-Bench Lite|GPT-4o|MaAS | -0.0783 | 1.390625 | 0.2002 | 0.1219 | 168 |
Jwalapuram | SWE-Bench Lite | GPT-4o | MaAS | B | CoT-SC (5) | Jwalapuram|SWE-Bench Lite|GPT-4o|MaAS | -0.0982 | 0.211527 | 0.2201 | 0.1219 | 168 |
Multi-agent versus single-agent contrasts from nine studies
Accuracy contrasts extracted from the tables of nine published evaluations of multi-agent language-model systems, with the single-agent baseline of each comparison tagged by rung of a baseline ladder.
Files
data/primary_contrasts.{parquet,csv}: 531 primary-baseline contrasts from 9 studies; 516 are multi-agent configurations (multi_agent_configurationis true). The other 15 rows are majority-voting arms of one study and a self-consistency row of another, which the analysis treats as baselines.data/baseline_tier_rows.{parquet,csv}: 832 rows; each multi-agent configuration against each single-agent baseline available for it.baseline_rungis A (single greedy pass, 358 rows), B (sampled baseline, 360 rows) or C (a named resource held equal, 114 rows). No row carries rung D (budget sweep).
Columns
primary_contrasts: contrast_id, study_id, benchmark, model, baseline (condition label of the single-agent arm), mas_condition, p_baseline, p_mas (accuracies as proportions), n_items, budget_metric, budget_value, source_file, multi_agent_configuration, accuracy_difference (p_mas minus p_baseline).
baseline_tier_rows: study, benchmark, model, ma_configuration, baseline_rung, baseline, key (study|benchmark|model|configuration), accuracy_difference, cost_ratio (Jwalapuram only), p_base, p_mas, n_items.
Provenance and caveats
Values were extracted from published source tables and verified against rendered source tables across a stratified audit sample of 40 contrasts (no discrepancy recorded). Rows within a study share baselines and are not independent. Source papers retain their original licenses; this compiled dataset is released under CC BY 4.0.
- Paper (arXiv): https://arxiv.org/abs/2611.17610
- HTML Version: https://arxiv.org/html/2611.17610v1
- Code & Replication Repository: https://github.com/mahmoudsajjadi/pace
- Downloads last month
- -