{"id":5818,"date":"2026-06-12T15:00:08","date_gmt":"2026-06-12T07:00:08","guid":{"rendered":"https:\/\/1ow.top\/index.php\/2026\/06\/12\/%e6%99%ba%e8%83%bd%e4%bd%93%e6%9c%80%e5%90%8e%e7%9a%84%e8%80%83%e8%af%95%ef%bc%8cfable-5%e7%ab%9f%e7%84%b6%e4%b8%8d%e6%95%8cgpt-5-5\/"},"modified":"2026-06-12T15:00:08","modified_gmt":"2026-06-12T07:00:08","slug":"%e6%99%ba%e8%83%bd%e4%bd%93%e6%9c%80%e5%90%8e%e7%9a%84%e8%80%83%e8%af%95%ef%bc%8cfable-5%e7%ab%9f%e7%84%b6%e4%b8%8d%e6%95%8cgpt-5-5","status":"publish","type":"post","link":"https:\/\/1ow.top\/index.php\/2026\/06\/12\/%e6%99%ba%e8%83%bd%e4%bd%93%e6%9c%80%e5%90%8e%e7%9a%84%e8%80%83%e8%af%95%ef%bc%8cfable-5%e7%ab%9f%e7%84%b6%e4%b8%8d%e6%95%8cgpt-5-5\/","title":{"rendered":"\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\uff0cFable 5\u7adf\u7136\u4e0d\u654cGPT 5.5"},"content":{"rendered":"
\"\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\uff0cFable<\/div><\/figure>\n
\"\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\uff0cFable<\/div><\/figure>\n
\"\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\uff0cFable<\/div><\/figure>\n
\"\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\uff0cFable<\/div><\/figure>\n

\u6ca1\u60f3\u5230\u6253\u8138\u6765\u5f97\u5982\u6b64\u4e4b\u5feb\uff01\uff01<\/p>\n

\u521a\u521a\uff0cUC\u4f2f\u514b\u5229\u653e\u51fa\u4e86\u4e00\u573a\u53f7\u79f0\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\u7684\u5168\u65b0\u57fa\u51c6\u6d4b\u8bd5\u3002<\/p>\n

\u5b83\u628a\u5f53\u4eca\u6700\u5f3a\u7684AI Agent\u4eec\u62c9\u5230\u8003\u573a\u4e0a\uff0c\u8ba9\u5b83\u4eec\u5e72\u771f\u6b63\u7684\u6d3b\u2014\u2014<\/p>\n

\u5728Siemens NX\u91cc\u5efa3D\u6a21\u578b\u3001\u5728Unreal Engine\u91cc\u642d\u6e38\u620f\u573a\u666f\u3001\u5728Adobe After Effects\u91cc\u505a\u7279\u6548\u5408\u6210\u3002<\/p>\n

\u7ed3\u679c\u6210\u7ee9\u4ee4\u4eba\u50bb\u773c\uff1a<\/p>\n

\u6700\u96be\u7684\u4e00\u6863\uff0c\u5f53\u4eca\u516c\u8ba4\u6700\u5f3a\u7684Claude Fable 5\u3001GPT 5.5\uff0c\u5168\u662f\u5927\u5199\u7684\u96f6\u86cb\u3002<\/p>\n

\u4f60\u8bf4\u96be\u5ea6\u7a0d\u5fae\u653e\u4f4e\u4e00\u70b9\u5462\uff1f\u5206\u6570\u5012\u662f\u6709\u4e86\uff0c\u4f46\u7ed3\u679c\u4e5f\u76f8\u5f53\u4ee4\u4eba\u610f\u5916\u2014\u2014<\/p>\n

GPT 5.5\u7adf\u7136\u8fd8\u5c0f\u80dc\u4e86Claude Fable 5\u3002<\/p>\n

\u6211\u6ca1\u542c\u9519\u5427\uff0cA\u5bb6\u521a\u53d1\u5e03\u7684\u6700\u5f3a\u6a21\u578bClaude Fable 5\uff0c\u88ab\u51e0\u4e2a\u6708\u524d\u7684GPT 5.5\u6253\u8d25\u4e86\uff1f\uff1f<\/p>\n

\u8981\u77e5\u9053\u5728\u6b64\u524d\u51e0\u4e4e\u6240\u6709\u4e3b\u6d41benchmark\u4e0a\uff0cFable 5\u5bf9GPT 5.5\u90fd\u662f\u78be\u538b\u7ea7\u522b\u7684\u5b58\u5728\u2014\u2014SWE-Bench Pro\u4e0a80.3%\u5bf958.6%\uff0cHumanity\u2019s Last Exam\u4e0a64.5%\u5bf952.2%\u3002<\/p>\n

\u4f46\u6362\u5230\u8fd9\u573a\u201c\u771f\u5e72\u6d3b\u201d\u7684\u8003\u8bd5\u91cc\uff0c\u5c40\u9762\u5374\u53cd\u4e86\u8fc7\u6765\u3002<\/p>\n

\u8fd9\u4e2a\u65b0\u57fa\u51c6\u53ebAgents\u2019 Last Exam\uff08ALE\uff09\uff0c\u80cc\u540e\u56e2\u961f\u6765\u5934\u4e0d\u5c0f\uff0c\u4e4b\u524dMMLU\u3001MATH\u3001CyberGym\u3001ExploitGym\u8fd9\u4e9b\u4f60\u8033\u719f\u80fd\u8be6\u7684\u57fa\u51c6\u90fd\u662f\u4ed6\u4eec\u63d0\u7684\u3002<\/p>\n

\u53d6\u8fd9\u4e2a\u540d\u4f30\u8ba1\u4e5f\u662f\u53c2\u8003\u4e4b\u524dScale AI\u90a3\u4e2a\u201cHumanity\u2019s Last Exam\u201d\uff08\u4eba\u7c7b\u6700\u540e\u7684\u8003\u8bd5\uff09\uff0c\u53ea\u4e0d\u8fc7\u8fd9\u6b21\u88ab\u8003\u7684\u4e0d\u662f\u4eba\u7c7b\u77e5\u8bc6\u7684\u6781\u9650\uff0c\u800c\u662fAI Agent\u5e72\u6d3b\u7684\u6781\u9650\u3002<\/p>\n

\u8be5\u8bf4\u4e0d\u8bf4\uff0c\u8fd9\u4e2a\u6d4b\u8bc4\u4e00\u51fa\u6765\uff0c\u4ee5\u524d\u5929\u5929\u558a\u7740\u201cAgent\u8981\u53d6\u4ee3\u4eba\u7c7b\u5de5\u4f5c\u201d\u7684\u4eba\uff0c\u8fd9\u4e0b\u662f\u771f\u5e72\u6c89\u9ed8\u4e86\u2026<\/p>\n

\u5148\u770b\u5b8c\u6574\u6392\u884c\u699c\u3002<\/p>\n

\u4ece\u6700\u6838\u5fc3\u7684\u4efb\u52a1\u901a\u8fc7\u7387\u6307\u6807\u6765\u770b\uff0cGPT 5.5\u76f4\u63a5\u5305\u63fd\u51a0\u4e9a\u519b\uff1a<\/p>\n

\u7b2c1\u540d\u662fGPT 5.5\u642d\u914dOpenAI\u81ea\u5bb6\u7684Codex\u6846\u67b6\uff0c\u901a\u8fc7\u738724.0%\u3002<\/p>\n

\u7b2c2\u540d\u8fd8\u662fGPT-5.5\uff0c\u53ea\u4e0d\u8fc7\u6362\u4e86ALE Claw\u6846\u67b6\uff0c\u901a\u8fc7\u738723.0%\u3002<\/p>\n

\uff08ALE Claw\u662f\u56e2\u961f\u81ea\u5df1\u5199\u7684\u4e00\u4e2abaseline Agent\uff0c\u8ddfCodex\u3001Claude Code\u3001Cursor CLI\u8fd9\u4e9b\u5546\u4e1a\u6846\u67b6\u5e76\u5217\u53c2\u8d5b\uff09<\/p>\n

\u76f4\u5230\u7b2c3\u540d\uff0c\u6211\u4eec\u624d\u770b\u5230Claude Fable 5\u7684\u8eab\u5f71\u2014\u2014\u642d\u914dClaude Code\uff0c\u62ff\u4e0b22.0%\u7684\u901a\u8fc7\u7387\u3002<\/p>\n

\u5f80\u4e0b\u770b\u66f4\u6709\u610f\u601d\u3002<\/p>\n

\u7b2c4\u3001\u7b2c5\u3001\u7b2c8\u540d\u5168\u662fGPT 5.5\uff0c\u53ea\u662f\u6362\u4e86\u4e0d\u540c\u7684\u6846\u67b6\u3002<\/p>\n

\u524d10\u540d\u91ccGPT 5.5\u51fa\u573a\u4e865\u6b21\uff0c\u52a0\u4e0a\u7b2c6\u540d\u7684GPT 5.4\uff0cOpenAI\u6a21\u578b\u76f4\u63a5\u5360\u4e866\u5e2d\u3002<\/p>\n

\u800cClaude\u5bb6\u65cf\u5462\uff1f<\/p>\n

Fable 5\u62ff\u4e86\u7b2c3\uff0cOpus 4.7\u7b2c9\uff0818.4%\uff09\uff0cOpus 4.8\u57ab\u5e95\u7b2c10\uff0815.8%\uff09\uff0c\u4e0d\u654c\u4e4b\u52bf\u4e00\u76ee\u4e86\u7136\u3002<\/p>\n

\u4e5f\u4e0d\u602aOpenAI\u7814\u7a76\u5458\u559c\u5e86\u53d1\u5e16\uff0c\u6b22\u6b22\u559c\u559c\u8fc7\u5927\u5e74\u4e86\uff1a<\/p>\n

\u800c\u5728\u6210\u7ee9\u4e4b\u5916\uff0c\u8fd9\u91cc\u8fd8\u6709\u8fd9\u6837\u51e0\u4e2a\u503c\u5f97\u7ec6\u54c1\u7684\u4fe1\u53f7\u3002<\/p>\n

\u4e00\u662f\u5929\u82b1\u677f\u4f4e\u5f97\u60ca\u4eba\u3002<\/p>\n

\u51a0\u519b\u901a\u8fc7\u7387\u624d24%\uff0c\u7efc\u5408\u5f97\u5206\u6700\u9ad8\u4e5f\u4e0d\u8fc745.8%\u3002<\/p>\n

\u610f\u601d\u662f\uff0c\u5c31\u7b97\u6309\u6700\u5bbd\u677e\u7684\u201c\u90e8\u5206\u5f97\u5206\u201d\u7b97\uff0c\u6700\u5f3a\u7684Agent\u4e5f\u53ea\u80fd\u62ff\u5230\u4e0d\u5230\u4e00\u534a\u7684\u5206\u3002<\/p>\n

\u800c\u8fd9\u4e9b\u9898\u5168\u90e8\u6765\u81ea\u771f\u4eba\u4e13\u5bb6\u5df2\u7ecf\u5b8c\u6210\u7684\u9879\u76ee\u2014\u2014\u4eba\u7c7b\u4e13\u5bb6\u7684\u5b8c\u6210\u7387\u7406\u8bba\u4e0a\u5c31\u662f100%\u3002<\/p>\n

\u4e8c\u662fClaude\u70e7\u94b1\u70e7\u5f97\u60ca\u4eba\u3002<\/p>\n

\u8fd9\u5f20\u699c\u5355\u65b0\u589e\u4e86\u4e00\u5217\u201cEstimated Total Cost\u201d\uff0c\u4e00\u4e0b\u5b50\u628a\u8d2b\u5bcc\u5dee\u8ddd\u62c9\u51fa\u6765\u4e86\uff1a<\/p>\n

Fable 5\u8dd1\u5b8c\u5168\u90e8\u4efb\u52a1\u82b1\u4e862315\u7f8e\u5143\uff0cOpus 4.8\u82b1\u4e861838\u7f8e\u5143\uff0cOpus 4.7\u4e5f\u89811144\u7f8e\u5143\u3002<\/p>\n

\u800cGPT-5.5\u8fd9\u8fb9\u5462\uff1f<\/p>\n

\u6700\u8d35\u7684Codex\u4e5f\u5c31566\u7f8e\u5143\uff0cCursor CLI\u53ea\u8981174\u7f8e\u5143\u3002<\/p>\n

\u7b49\u4e8e\u8bf4\uff0cFable 5\u82b1\u4e86Codex\u56db\u500d\u591a\u7684\u94b1\uff0c\u6210\u7ee9\u8fd8\u4f4e\u4e86\u4e24\u4e2a\u767e\u5206\u70b9\u3002<\/p>\n

\u4e09\u662f\u6548\u7387\u5dee\u8ddd\u540c\u6837\u89e6\u76ee\u3002<\/p>\n

Ale Claw\u8dd1\u5b8c\u5168\u90e8\u4efb\u52a1\u82b1\u4e8647\u5c0f\u65f620\u5206\u949f\uff0cCursor CLI\u53ea\u82b1\u4e8667\u5c0f\u65f6\u3002<\/p>\n

\u800cOpus 4.8\u5462\uff1f451\u5c0f\u65f6\u2014\u2014\u5c06\u8fd119\u5929\u3002<\/p>\n

\u5e72\u7684\u6d3b\u6700\u5c11\uff0c\u82b1\u7684\u65f6\u95f4\u6700\u957f\uff0c\u6536\u7684\u94b1\u6700\u591a\uff08\u5c45\u7136\u771f\u6709\u6a21\u578b\u80fd\u540c\u65f6\u505a\u5230\uff1f\uff09<\/p>\n

\u5f53\u7136\u5982\u679c\u53ea\u770bClaude Fable 5\u3001GPT 5.5\u8fd9\u4e24\u4e2a\u6700\u9876\u7684\uff0cGPT 5.5\u7684\u65f6\u95f4\u4f18\u52bf\u4f9d\u65e7\u660e\u663e\u3002<\/p>\n

\u800c\u6700\u624e\u773c\u7684\u6570\u5b57\uff0c\u8fd8\u662f\u90a3\u4e2a\u96f6\u3002<\/p>\n

ALE\u628a\u4efb\u52a1\u5206\u6210\u4e86\u4e09\u4e2a\u96be\u5ea6\u6863\uff1a<\/p>\n

\u5728\u6700\u96be\u8fd9\u4e00\u6863\uff0c\u6240\u6709\u4e3b\u6d41\u914d\u7f6e\u7684\u5e73\u5747\u901a\u8fc7\u7387\u53ea\u67092.6%\uff0c\u5305\u62ecGPT 5.5\u548cFable 5\u5728\u5185\u7684\u5927\u591a\u6570\u6a21\u578b\u76f4\u63a5\u5403\u4e86\u96f6\u86cb\u3002<\/p>\n

\u6240\u4ee5\u8fd9\u5f20\u6210\u7ee9\u5355\u7684\u6838\u5fc3\u4fe1\u606f\u5f88\u7b80\u5355\uff1a\u522b\u770b\u5e73\u65f6\u8003\u8bd5\u6210\u7ee9\u597d\uff0c\u4e00\u5230\u771f\u5e72\u6d3b\u5168\u9732\u9985\u4e86\u3002<\/p>\n

\u7b54\u9898\u5b66\u9738\u2260\u5e72\u6d3b\u80fd\u624b\uff0c\u8fd9\u8bdd\u5728AI\u4e16\u754c\u4e5f\u4e00\u6837\u9002\u7528\u3002<\/p>\n

\u8981\u7406\u89e3ALE\u4e3a\u4ec0\u4e48\u80fd\u628a\u8fd9\u5e2e\u201c\u5b66\u9738\u201d\u6253\u56de\u539f\u5f62\uff0c\u5f97\u5148\u770b\u5b83\u8ddf\u4ee5\u524d\u7684\u8003\u8bd5\u6709\u4ec0\u4e48\u4e0d\u4e00\u6837\u3002<\/p>\n

\u4e4b\u524d\u7684Humanity\u2019s Last Exam\uff08HLE\uff09\u662f2025\u5e74\u521d\u7531Dan Hendrycks\u548cScale AI\u641e\u51fa\u6765\u7684\uff0c2500\u9053\u8de8\u5b66\u79d1\u96be\u9898\uff0c\u672c\u8d28\u4e0a\u8fd8\u662f\u95ed\u5377\u7b54\u9898\u2014\u2014<\/p>\n

\u7ed9\u4f60\u4e00\u4e2a\u95ee\u9898\uff0c\u4f60\u7ed9\u6211\u4e00\u4e2a\u7b54\u6848\uff0c\u518d\u96be\u4e5f\u662f\u9759\u6001\u7684\u77e5\u8bc6\u68c0\u7d22\u3002<\/p>\n

\u800cALE\u5b8c\u5168\u4e0d\u540c\uff0c\u5b83\u8003\u4f60\u201c\u80fd\u5e72\u4ec0\u4e48\u201d\u3002<\/p>\n

\u6838\u5fc3\u4f5c\u8005Yiyou Sun\u5728\u8bf4\u5f97\u5f88\u76f4\u767d\uff1a<\/p>\n

AI\u667a\u80fd\u4f53\u5c06\u57282026-2027\u5e74\u8d85\u8d8a\u4eba\u7c7b\u5b8c\u6210\u51e0\u4e4e\u6240\u6709\u5de5\u4f5c\u2014\u2014\u8fd9\u4e2a\u9884\u6d4b\u5230\u5904\u90fd\u662f\u3002\u6240\u4ee5\u6211\u4eec\u9020\u4e86\u8fd9\u573a\u8003\u8bd5\u6765\u9a8c\u8bc1\u8fd9\u4e2a\u8bf4\u6cd5\u3002<\/p>\n

ALE\u7684\u6bcf\u9053\u9898\u90fd\u6765\u81ea\u4e00\u4e2a\u771f\u4eba\u4e13\u5bb6\u5df2\u7ecf\u5b8c\u6210\u7684\u9879\u76ee\uff0c\u8986\u76d655\u4e2a\u884c\u4e1a\u5b50\u9886\u57df\uff0c\u5305\u62ec\u91cf\u5316\u4ea4\u6613\u3001\u57fa\u56e0\u7ec4\u5206\u6790\u3001\u822a\u7a7a\u822a\u5929\u5de5\u7a0b\u3001\u5efa\u7b51\u8bbe\u8ba1\u3001\u8111\u6210\u50cf\u3001\u52a8\u753b\u7279\u6548\u3001\u6cd5\u5f8b\u7814\u7a76\u2026\u2026<\/p>\n

\u6574\u4e2a\u4f53\u7cfb\u951a\u5b9a\u7684\u662f\u7f8e\u56fd\u8054\u90a6\u804c\u4e1a\u5206\u7c7b\u6807\u51c6\uff08ONET\uff09*\uff0c\u8bf4\u767d\u4e86\u5c31\u662f\u6309\u201c\u771f\u5b9e\u52b3\u52a8\u529b\u5e02\u573a\u201d\u6765\u51fa\u9898\u3002<\/p>\n

\u53c2\u4e0e\u51fa\u9898\u7684\u9635\u5bb9\u4e5f\u591f\u8c6a\u534e\uff1a<\/p>\n

300\u591a\u4f4d\u9886\u57df\u4e13\u5bb6\u6765\u81ea100\u591a\u5bb6\u673a\u6784\uff0c\u5b66\u672f\u4fa7\u6709MIT\u3001Harvard\u3001Stanford\u3001Oxford\u3001Caltech\u3001ETH Zurich\uff0c\u4ea7\u4e1a\u4fa7\u6709Goldman Sachs\u3001JPMorgan\u3001Meta\u3001Amazon\u3001Adobe\u3001Oracle\u3002<\/p>\n

Snorkel AI\u901a\u8fc7Open Benchmarks Grants\u9879\u76ee\u63d0\u4f9b\u4e86\u8d44\u91d1\u652f\u6301\u3002<\/p>\n

\u8003\u8bd5\u5f62\u5f0f\u4e5f\u4e0d\u662f\u6253\u5b57\u56de\u7b54\u95ee\u9898\uff0c\u800c\u662f\u76f4\u63a5\u64cd\u4f5c\u7535\u8111\u3002<\/p>\n

ALE\u7528\u7684\u662f\u6240\u8c13GCUA\u6846\u67b6\uff08Generalist Computer-Use Agent\uff0c\u901a\u7528\u8ba1\u7b97\u673a\u4f7f\u7528\u4ee3\u7406\uff09\uff0c\u7ed9Agent\u5b8c\u6574\u7684GUI\u548c\u547d\u4ee4\u884c\u6743\u9650\u2014\u2014<\/p>\n

\u9f20\u6807\u70b9\u51fb\u3001\u952e\u76d8\u6253\u5b57\u3001\u5199\u811a\u672c\u3001\u6d4f\u89c8\u7f51\u9875\uff0c\u4eba\u7c7b\u80fd\u5728\u7535\u8111\u4e0a\u5e72\u7684\u5b83\u90fd\u80fd\u5e72\u3002<\/p>\n

\u4e0d\u9650\u65b9\u6cd5\uff0c\u53ea\u770b\u7ed3\u679c\u3002<\/p>\n

\u4ea4\u51fa\u6765\u7684\u201c\u4f5c\u4e1a\u201d\u7531\u786e\u5b9a\u6027\u4ee3\u7801\u81ea\u52a8\u8bc4\u5206\u3002<\/p>\n

No vibes. No human judges. Fully reproducible.\uff08\u4e0d\u9760\u611f\u89c9\uff0c\u4e0d\u9760\u4eba\u7c7b\u88c1\u5224\uff0c\u5b8c\u5168\u53ef\u590d\u73b0\uff09<\/p>\n

\u8fd9\u5c31\u5835\u4f4f\u4e86\u4e4b\u524d\u5f88\u591abenchmark\u7684\u4e00\u4e2a\u8001\u6bdb\u75c5\uff1a\u8bc4\u5206\u5668\u672c\u8eab\u5c31\u80fd\u88ab\u9a97\u3002<\/p>\n

\u6b64\u5916\uff0cALE\u5728\u9632\u4f5c\u5f0a\u4e0a\u8fd8\u6709\u4e00\u4e2a\u72e0\u62db\u2014\u2014<\/p>\n

\u53ea\u516c\u5f00\u7ea610%\u7684\u9898\u76ee\uff08\u7ea6150\u9053\uff09\uff0c\u5269\u4e0b1300\u591a\u9053\u4e25\u683c\u4fdd\u5bc6\u3002<\/p>\n

\u516c\u5f00\u9898\u548c\u79c1\u5bc6\u9898\u5b9a\u671f\u6eda\u52a8\u8f6e\u6362\uff0c\u786e\u4fdd\u4e0d\u4f1a\u6709\u6a21\u578b\u56e0\u4e3a\u201c\u80cc\u9898\u201d\u800c\u62ff\u9ad8\u5206\u3002<\/p>\n

\u8fd9\u5728\u5f53\u524dbenchmark\u6570\u636e\u6c61\u67d3\u6cdb\u6ee5\u7684\u80cc\u666f\u4e0b\uff0c\u7b97\u662f\u4e00\u4e2a\u76f8\u5f53\u5de7\u5999\u7684\u8bbe\u8ba1\u3002<\/p>\n

\u6574\u4f53\u800c\u8a00\uff0c\u8ddf\u73b0\u6709\u7684Agent\u57fa\u51c6\u6d4b\u8bd5\u6bd4\uff0cALE\u7684\u5b9a\u4f4d\u975e\u5e38\u660e\u786e\u3002<\/p>\n

\u56e2\u961f\u6210\u5458\u4e4b\u4e00\u7684Dawn Song\u4e13\u95e8\u62c9\u4e86\u4e00\u7ec4\u5bf9\u6bd4\uff1a<\/p>\n

\u4e00\u8a00\u4ee5\u853d\u4e4b\uff0c\u5176\u4ed6\u8003\u8bd5\u5df2\u7ecf\u5feb\u88ab\u505a\u7a7f\u4e86\uff0c\u800cALE\u8fd8\u8fdc\u5f97\u5f88\u3002<\/p>\n

\u8fd9\u5c31\u662fALE\u51ed\u4ec0\u4e48\u6562\u81ea\u79f0\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\u7684\u7406\u7531\u3002<\/p>\n

\u503c\u5f97\u4e00\u63d0\u7684\u662f\uff0cDawn Song\u8fd8\u5206\u4eab\u4e86\u4e24\u4e2a\u6709\u8da3\u7684\u89c2\u5bdf\uff1a<\/p>\n

\u4e00\u4e2a\u662f\uff0cAgent\u4f1a\u5728\u6ca1\u6709\u771f\u6b63\u9a8c\u8bc1\u5de5\u4f5c\u6210\u679c\u7684\u60c5\u51b5\u4e0b\u5ba3\u5e03\u5b8c\u6210\uff0c\u8fd9\u662fAgent\u4eec\u6700\u5178\u578b\u7684\u5931\u8d25\u6a21\u5f0f\u3002<\/p>\n

\u5f88\u591a\u65f6\u5019\uff0c\u867d\u7136\u5b83\u4eec\u8bf4\u4e86\u201cDone. All checks pass.\u201d\uff08\u641e\u5b9a\u4e86\uff0c\u6240\u6709\u68c0\u67e5\u90fd\u901a\u8fc7\u4e86\uff09<\/p>\n

\u4f46\u5b9e\u9645\u4ea7\u51fa\u53ef\u80fd\u7f3a\u5c11\u5fc5\u8981\u6587\u4ef6\u3001\u6570\u5b57\u7b97\u9519\u3001\u5173\u952e\u5b57\u6bb5\u9057\u6f0f\u3001\u6216\u8005\u76f4\u63a5\u8fdd\u53cd\u4e86\u4efb\u52a1\u8bf4\u660e\u4e2d\u7684\u660e\u786e\u7ea6\u675f\u3002<\/p>\n

\u7b49\u4e8e\u662f\uff0c\u6d3b\u6ca1\u5e72\u5b8c\uff0c\u5634\u5148\u8bf4\u5b8c\u4e86\u3002<\/p>\n

\u53e6\u4e00\u4e2a\u662f\u5f88\u591a\u4eba\u7591\u60d1\u7684\uff0c\u4e3a\u5565Fable 5\u8fd9\u4e48\u62c9\u80ef\uff1fDawn Song\u7ed9\u51fa\u7684\u56de\u7b54\u662f\uff1a<\/p>\n

\u4e0d\u5b58\u5728\u201c\u4e07\u80fd\u51a0\u519b\u201d\u8fd9\u56de\u4e8b\u3002<\/p>\n

\u6bcf\u4e2a\u524d\u6cbf\u6a21\u578b\u90fd\u6709\u64c5\u957f\u7684\u9886\u57df\u548c\u62c9\u80ef\u7684\u9886\u57df\uff0cALE\u8986\u76d655\u4e2a\u884c\u4e1a\u30011500+\u9053\u9898\uff0c\u6700\u7ec8\u5f97\u5206\u662f\u6240\u6709\u9886\u57df\u7684\u5e73\u5747\u503c\uff0c\u5f88\u591a\u6a21\u578b\u7684\u603b\u5206\u56e0\u6b64\u6324\u5728\u4e00\u8d77\u3002\u771f\u6b63\u6709\u4ef7\u503c\u7684\u4fe1\u53f7\u4e0d\u5728\u603b\u5206\uff0c\u800c\u5728\u4e0d\u540c\u6a21\u578b\u5728\u4e0d\u540c\u9886\u57df\u7684\u8868\u73b0\u5dee\u5f02\u2014\u2014\u5728\u540c\u4e00\u9053\u9898\u4e0a\uff0c\u4e0d\u540c\u6a21\u578b\u5f80\u5f80\u56e0\u4e3a\u5b8c\u5168\u4e0d\u540c\u7684\u539f\u56e0\u800c\u5931\u8d25\u3002<\/p>\n

\u5f53\u7136\u4e5f\u6709\u53ef\u80fd\u662fFable 5\u5077\u5077\u201c\u964d\u667a\u201d\u4e86\u3002<\/p>\n

\u603b\u699c\u91cc\uff0cFable 5\u65c1\u8fb9\u6807\u9ec4\u4e86\u4e00\u53e5\u201cmay be down-tuned\u201d\uff08\u53ef\u80fd\u88ab\u964d\u7ea7\uff09\uff0c\u8fd9\u8bf4\u7684\u662fFable 5\u7684\u4e00\u4e2a\u5df2\u77e5\u95ee\u9898\u2014\u2014<\/p>\n

\u5b83\u5e95\u5c42\u662fMythos\u6a21\u578b\u52a0\u5b89\u5168\u5206\u7c7b\u5668\uff0c\u9047\u5230\u7f51\u7edc\u5b89\u5168\u3001\u751f\u7269\u533b\u5b66\u7b49\u654f\u611f\u9886\u57df\u7684\u4efb\u52a1\u65f6\uff0c\u4f1a\u88ab\u9759\u9ed8\u5207\u6362\u5230\u80fd\u529b\u66f4\u5f31\u7684Opus 4.8\u3002<\/p>\n

\u5728ALE\u8fd9\u79cd\u8986\u76d655\u4e2a\u884c\u4e1a\u7684\u8003\u8bd5\u4e2d\uff0c\u7b49\u4e8e\u8fd9\u90e8\u5206\u79d1\u76ee\u76f4\u63a5\u6d3e\u4e86\u66ff\u8003\uff0c\u800c\u4e14\u6d3e\u7684\u8fd8\u662f\u201c\u5954\u6ce2\u513f\u705e\u201d\u8fd9\u79cd\u89d2\u8272\u3002<\/p>\n

\u5f53\u7136\uff0c\u6709\u6ca1\u6709\u53ef\u80fdClaude Fable 5\u7684\u6210\u7ee9\u672c\u8eab\u5c31\u6709\u95ee\u9898\u5462\uff1f<\/p>\n

\u4e0d\u597d\u8bf4\uff0c\u4f46\u4e00\u6869\u516b\u5366\u663e\u793a\uff0cClaude\u6709\u201c\u524d\u79d1\u201d\u3002<\/p>\n

5\u6708\u5e95\uff0c\u521d\u521b\u516c\u53f8Datacurve\u53d1\u5e03\u4e86\u4e00\u4e2a\u53ebDeepSWE\u7684\u65b0benchmark\uff0c\u987a\u624b\u63ed\u4e86\u4e00\u4e2a\u5927\u5e95\u2014\u2014<\/p>\n

SWE-Bench Pro\u7684Docker\u5bb9\u5668\u91cc\u9644\u5e26\u4e86\u4ee3\u7801\u4ed3\u5e93\u7684\u5b8c\u6574git\u5386\u53f2\uff0c\u6b63\u786e\u7b54\u6848\u5c31\u8eba\u5728\u6587\u4ef6\u7cfb\u7edf\u91cc\u3002<\/p>\n

\u5927\u591a\u6570\u6a21\u578b\u4f1a\u65e0\u89c6\u5b83\uff0c\u4f46\u53ea\u6709Claude\u4e0d\u4f1a\u3002<\/p>\n

\u5b83\u4f1a\u4e3b\u52a8\u68c0\u67e5\u4ed3\u5e93\u7684git\u5386\u53f2\uff0c\u4ece\u5386\u53f2\u63d0\u4ea4\u4e2d\u5bfb\u627e\u4e0e\u4efb\u52a1\u5bf9\u5e94\u7684\u4fee\u590d\u65b9\u6848\uff0c\u5e76\u636e\u6b64\u6062\u590d\u6b63\u786e\u8865\u4e01\u3002<\/p>\n

\u636e\u79f0Opus 4.7\u7ea618%\u7684\u901a\u8fc7\u6210\u7ee9\u662f\u8fd9\u4e48\u62ff\u7684\uff0cOpus 4.6\u66f4\u5938\u5f20\uff0c\u7ea625%\u3002<\/p>\n

\u800cGPT 5.4\u548cGPT5.5\u8fd9\u8fb9\u5462\uff1f\u5b8c\u5168\u6ca1\u6709\u8fd9\u79cd\u884c\u4e3a\u3002Datacurve\u7684\u63aa\u8f9e\u5f88\u5916\u4ea4\uff1a<\/p>\n

\u8fd9\u4e2abenchmark\u8ba9\u8fd9\u79cd\u884c\u4e3a\u6210\u4e3a\u53ef\u80fd\uff0c\u4f46Claude\u662f\u552f\u4e00\u6301\u7eed\u8fd9\u4e48\u505a\u7684\u5bb6\u65cf\u3002<\/p>\n

\u79d1\u6280\u5a92\u4f53VentureBeat\u7684\u8bc4\u4ef7\u5012\u5f88\u66a7\u6627\uff1a<\/p>\n

\u8fd9\u8bf4\u660eClaude\u201c\u73af\u5883\u611f\u77e5\u80fd\u529b\u201d\u5f88\u5f3a\uff0c\u975e\u5e38\u64c5\u957f\u63a2\u7d22\u5468\u56f4\u73af\u5883\u5e76\u5229\u7528\u53ef\u7528\u8d44\u6e90\u3002\u7b97\u201c\u4f5c\u5f0a\u201d\u8fd8\u662f\u201c\u673a\u7075\u201d\uff0c\u53d6\u51b3\u4e8e\u4f60\u7684\u7acb\u573a\u3002<\/p>\n

\u4f46\u752d\u7ba1\u600e\u4e48\u770b\uff0cALE\u663e\u7136\u5438\u53d6\u4e86\u6559\u8bad\u2014\u2014<\/p>\n

\u76f4\u63a5\u628a\u8003\u573a\u4ece\u547d\u4ee4\u884c\u642c\u5230\u4e86GUI\u684c\u9762\u64cd\u4f5c\uff0c\u8ba9\u4f60\u6ca1\u6709git\u5386\u53f2\u53ef\u4ee5\u5077\u770b\u3002<\/p>\n

\u8bc4\u6d4bAI\u7684\u8003\u573a\uff0c\u6b63\u5728\u88abAI\u81ea\u5df1\u5012\u903c\u7740\u5347\u7ea7\uff0c\u4e5f\u7b97\u5f88\u7cbe\u5f69\u4e86\u3002<\/p>\n

\u5b8c\u6574\u6d4b\u8bc4\u5730\u5740\uff1ahttps:\/\/agents-last-exam.org\/leaderboard\u9879\u76ee\u4e3b\u9875\uff1ahttps:\/\/agents-last-exam.org\/GitHub\uff1ahttps:\/\/github.com\/rdi-berkeley\/agents-last-exam<\/p>\n

\n

\u7248\u6743\u58f0\u660e\uff1a<\/strong><\/p>\n

\u672c\u535a\u5ba2\u90e8\u5206\u5185\u5bb9\u4e3a\u8f6c\u8f7d\u6587\u7ae0\uff0c\u65e8\u5728\u5206\u4eab\u6709\u4ef7\u503c\u7684\u4fe1\u606f\uff0c\u7248\u6743\u5f52\u539f\u4f5c\u8005\u6240\u6709\u3002<\/p>\n

\u8f6c\u8f7d\u4ec5\u4e3a\u4e2a\u4eba\u5b66\u4e60\u4e0e\u4ea4\u6d41\u76ee\u7684\uff0c\u4e0d\u5bf9\u6587\u7ae0\u89c2\u70b9\u8d1f\u8d23\uff0c\u4ea6\u4e0d\u7528\u4e8e\u4efb\u4f55\u5546\u4e1a\u7528\u9014\u3002<\/p>\n

\u5982\u6d89\u53ca\u7248\u6743\u95ee\u9898\uff0c\u8bf7\u8054\u7cfb\u672c\u4eba\u5220\u9664\u3002<\/p>\n<\/blockquote>\n","protected":false},"excerpt":{"rendered":"

\u6ca1\u60f3\u5230\u6253\u8138\u6765\u5f97\u5982\u6b64\u4e4b\u5feb\uff01\uff01<\/p>\n

\u521a\u521a\uff0cUC\u4f2f\u514b\u5229\u653e\u51fa\u4e86\u4e00\u573a\u53f7\u79f0\u201c\u667a\u80fd\u4f53\u6700\u540e\u7684\u8003\u8bd5\u201d\u7684\u5168\u65b0\u57fa\u51c6\u6d4b\u8bd5\u3002<\/p>\n

\u5b83\u628a\u5f53\u4eca\u6700\u5f3a\u7684AI Agent\u4eec\u62c9\u5230\u8003\u573a\u4e0a\uff0c\u8ba9\u5b83\u4eec\u5e72\u771f\u6b63\u7684\u6d3b\u2014\u2014<\/p>\n

\u5728Siemens NX\u91cc\u5efa3D\u6a21\u578b\u3001\u5728Unreal Engine\u91cc\u642d\u6e38\u620f\u573a\u666f\u3001\u5728Adobe After Effects\u91cc\u505a\u7279\u6548\u5408\u6210\u3002<\/p>\n

\u7ed3\u679c\u6210\u7ee9\u4ee4\u4eba\u50bb\u773c\uff1a<\/p>\n

\u6700\u96be\u7684\u4e00\u6863\uff0c\u5f53\u4eca\u516c\u8ba4\u6700\u5f3a\u7684Claude Fable 5\u3001GPT 5.5\uff0c\u5168<\/p>\n","protected":false},"author":1,"featured_media":5814,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-5818","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-xwzx"],"_links":{"self":[{"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/posts\/5818","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/comments?post=5818"}],"version-history":[{"count":0,"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/posts\/5818\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/media\/5814"}],"wp:attachment":[{"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/media?parent=5818"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/categories?post=5818"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/1ow.top\/index.php\/wp-json\/wp\/v2\/tags?post=5818"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}