Hugging Face's logo

Spaces:
SamsungResearch
/
TRUEBench
Running

App Files Community
4
j_yoon.song commited on
Commit
7d35c34
·
1 Parent(s): 9875a12

fix typo

Browse files
Files changed (2) hide show
  1. src/data/open/stats.csv +1 -1
  2. src/data/open/stats_lang.csv +1 -1
src/data/open/stats.csv CHANGED
@@ -6,7 +6,7 @@
6
  "Claude 4 Opus (20250514) (think)" "https://www.anthropic.com/claude/opus" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.29" "60.75" "59.69" "73.31" "69.83" "78.74" "53.72" "55.71" "65.48" "65.45" "48.99"
7
  "Claude 4.1 Opus (20250805) (think)" "https://www.anthropic.com/claude/opus" "version: 20250805" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.24" "61.25" "60.0" "78.49" "72.73" "77.01" "56.2" "57.14" "61.9" "62.64" "46.98"
8
  "GPT-5 mini (Reasoning: medium)" "https://platform.openai.com/docs/models/gpt-5-mini" "Reasoning: medium" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "62.56" "68.0" "62.5" "74.9" "76.86" "55.17" "47.93" "44.29" "74.6" "56.18" "45.3"
9
- "Gemini 3 Pro Preview (Thinking Level: High)" "" "" "Gemini" "1930.5" "378.0" "" "27.89457416534424" "" "" "Open" "Think" "On" "62.48" "59.5" "64.38" "76.49" "78.93" "70.69" "39.67" "65.71" "61.51" "58.15" "48.99"
10
  "Claude 4 Sonnet (20250514) (think)" "https://www.anthropic.com/claude/sonnet" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "61.8" "58.0" "58.44" "76.49" "67.77" "79.31" "57.02" "44.29" "65.08" "62.92" "44.97"
11
  "o3" "https://platform.openai.com/docs/models/o3" "" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "60.91" "68.75" "60.0" "73.31" "79.34" "54.02" "34.71" "64.29" "60.71" "55.06" "46.98"
12
  "Gemini 2.5 Pro" "https://deepmind.google/models/gemini/pro/" "" "Gemini" "" "" "" "" "" "" "Proprietary" "Think" "On" "59.34" "54.0" "60.94" "78.88" "73.14" "63.22" "17.36" "52.86" "67.86" "53.93" "52.68"
 
6
  "Claude 4 Opus (20250514) (think)" "https://www.anthropic.com/claude/opus" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.29" "60.75" "59.69" "73.31" "69.83" "78.74" "53.72" "55.71" "65.48" "65.45" "48.99"
7
  "Claude 4.1 Opus (20250805) (think)" "https://www.anthropic.com/claude/opus" "version: 20250805" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.24" "61.25" "60.0" "78.49" "72.73" "77.01" "56.2" "57.14" "61.9" "62.64" "46.98"
8
  "GPT-5 mini (Reasoning: medium)" "https://platform.openai.com/docs/models/gpt-5-mini" "Reasoning: medium" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "62.56" "68.0" "62.5" "74.9" "76.86" "55.17" "47.93" "44.29" "74.6" "56.18" "45.3"
9
+ "Gemini 3 Pro Preview (Thinking Level: High)" "" "" "Gemini" "1930.5" "378.0" "" "27.89457416534424" "" "" "Proprietary" "Think" "On" "62.48" "59.5" "64.38" "76.49" "78.93" "70.69" "39.67" "65.71" "61.51" "58.15" "48.99"
10
  "Claude 4 Sonnet (20250514) (think)" "https://www.anthropic.com/claude/sonnet" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "61.8" "58.0" "58.44" "76.49" "67.77" "79.31" "57.02" "44.29" "65.08" "62.92" "44.97"
11
  "o3" "https://platform.openai.com/docs/models/o3" "" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "60.91" "68.75" "60.0" "73.31" "79.34" "54.02" "34.71" "64.29" "60.71" "55.06" "46.98"
12
  "Gemini 2.5 Pro" "https://deepmind.google/models/gemini/pro/" "" "Gemini" "" "" "" "" "" "" "Proprietary" "Think" "On" "59.34" "54.0" "60.94" "78.88" "73.14" "63.22" "17.36" "52.86" "67.86" "53.93" "52.68"
src/data/open/stats_lang.csv CHANGED
@@ -6,7 +6,7 @@
6
  "Claude 4 Opus (20250514) (think)" "https://www.anthropic.com/claude/opus" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.29" "57.5" "62.5" "64.46" "62.8" "59.44" "65.19" "65.92" "60.54" "65.22" "65.57" "65.17" "72.56"
7
  "Claude 4.1 Opus (20250805) (think)" "https://www.anthropic.com/claude/opus" "version: 20250805" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.24" "58.33" "61.39" "60.84" "64.02" "61.67" "66.85" "68.16" "61.08" "65.76" "66.67" "65.73" "65.24"
8
  "GPT-5 mini (Reasoning: medium)" "https://platform.openai.com/docs/models/gpt-5-mini" "Reasoning: medium" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "62.56" "57.5" "56.39" "62.65" "62.2" "63.89" "60.22" "66.48" "67.03" "70.11" "67.76" "66.29" "60.98"
9
- "Gemini 3 Pro Preview (Thinking Level: High)" "" "" "Gemini" "1930.5" "378.0" "" "27.89457416534424" "" "" "Open" "Think" "On" "62.48" "59.44" "60.56" "60.24" "62.2" "61.67" "65.19" "63.13" "64.32" "65.76" "65.57" "64.04" "62.2"
10
  "Claude 4 Sonnet (20250514) (think)" "https://www.anthropic.com/claude/sonnet" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "61.8" "54.17" "59.17" "63.86" "64.63" "59.44" "61.33" "64.8" "62.16" "65.22" "67.21" "66.29" "64.02"
11
  "o3" "https://platform.openai.com/docs/models/o3" "" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "60.91" "57.5" "59.17" "61.45" "58.54" "61.11" "64.09" "60.89" "62.16" "63.59" "65.03" "54.49" "68.29"
12
  "Gemini 2.5 Pro" "https://deepmind.google/models/gemini/pro/" "" "Gemini" "" "" "" "" "" "" "Proprietary" "Think" "On" "59.34" "53.61" "57.78" "59.04" "57.93" "57.22" "56.91" "60.89" "63.24" "67.93" "62.3" "61.24" "60.98"
 
6
  "Claude 4 Opus (20250514) (think)" "https://www.anthropic.com/claude/opus" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.29" "57.5" "62.5" "64.46" "62.8" "59.44" "65.19" "65.92" "60.54" "65.22" "65.57" "65.17" "72.56"
7
  "Claude 4.1 Opus (20250805) (think)" "https://www.anthropic.com/claude/opus" "version: 20250805" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "63.24" "58.33" "61.39" "60.84" "64.02" "61.67" "66.85" "68.16" "61.08" "65.76" "66.67" "65.73" "65.24"
8
  "GPT-5 mini (Reasoning: medium)" "https://platform.openai.com/docs/models/gpt-5-mini" "Reasoning: medium" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "62.56" "57.5" "56.39" "62.65" "62.2" "63.89" "60.22" "66.48" "67.03" "70.11" "67.76" "66.29" "60.98"
9
+ "Gemini 3 Pro Preview (Thinking Level: High)" "" "" "Gemini" "1930.5" "378.0" "" "27.89457416534424" "" "" "Proprietary" "Think" "On" "62.48" "59.44" "60.56" "60.24" "62.2" "61.67" "65.19" "63.13" "64.32" "65.76" "65.57" "64.04" "62.2"
10
  "Claude 4 Sonnet (20250514) (think)" "https://www.anthropic.com/claude/sonnet" "version: 20250514" "Claude" "" "" "" "" "" "" "Proprietary" "Hybrid" "On" "61.8" "54.17" "59.17" "63.86" "64.63" "59.44" "61.33" "64.8" "62.16" "65.22" "67.21" "66.29" "64.02"
11
  "o3" "https://platform.openai.com/docs/models/o3" "" "GPT" "" "" "" "" "" "" "Proprietary" "Think" "On" "60.91" "57.5" "59.17" "61.45" "58.54" "61.11" "64.09" "60.89" "62.16" "63.59" "65.03" "54.49" "68.29"
12
  "Gemini 2.5 Pro" "https://deepmind.google/models/gemini/pro/" "" "Gemini" "" "" "" "" "" "" "Proprietary" "Think" "On" "59.34" "53.61" "57.78" "59.04" "57.93" "57.22" "56.91" "60.89" "63.24" "67.93" "62.3" "61.24" "60.98"