{"id":798788,"date":"2026-06-13T14:15:49","date_gmt":"2026-06-13T06:15:49","guid":{"rendered":"https:\/\/www.alphaseek.cn\/?p=798788"},"modified":"2026-06-13T14:15:49","modified_gmt":"2026-06-13T06:15:49","slug":"%e4%ba%a4%e7%99%bd%e5%8d%b7%e4%b9%9f%e6%8e%92%e7%ac%ac%e4%b8%80%ef%bc%9ffable-5%e4%ba%8c%e7%99%be%e9%a2%98%e5%85%a8%e9%83%a8%e6%8b%92%e7%ad%94%ef%bc%8c%e5%8d%b4%e7%99%bb%e9%a1%b6%e6%9c%80%e4%b8%a5ai","status":"publish","type":"post","link":"https:\/\/www.alphaseek.cn\/?p=798788","title":{"rendered":"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable 5\u4e8c\u767e\u9898\u5168\u90e8\u62d2\u7b54\uff0c\u5374\u767b\u9876\u6700\u4e25AI\u7f16\u7a0b\u57fa\u51c6"},"content":{"rendered":"

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u7f16\u8f91\uff5cPanda<\/p>\n

\u592a\u79bb\u8c31\u4e86\uff01<\/p>\n

\u662f\u7684\uff0c\u672c\u6587\u7684\u4e3b\u89d2\u8fd8\u662f\u524d\u4e9b\u5929\u521a\u53d1\u5e03\u4e86 Claude Fable 5 \u7684 Anthropic\u3002<\/p>\n

Fable 5 \u53d1\u5e03\u65f6\u7684\u6392\u9762\u662f\u771f\u7684\u8db3\u3002SWE-Bench Pro \u5f97\u5206 80.3%\uff0c\u628a\u7b2c\u4e8c\u540d\u7529\u51fa 11 \u4e2a\u767e\u5206\u70b9\uff1bAndrej Karpathy \u76f4\u63a5\u558a\u51fa\u300cdeserves a major version bump\u300d\uff1bStripe \u62ff\u5b83\u5728 5000 \u4e07\u884c Ruby \u4ee3\u7801\u5e93\u91cc\u8dd1\u4e86\u4e00\u6574\u5929\u7684\u8fc1\u79fb\uff0c\u9876\u4e0a\u4e86\u539f\u672c\u9700\u8981\u6574\u4e2a\u56e2\u961f\u4e24\u4e2a\u6708\u624d\u80fd\u5b8c\u6210\u7684\u5de5\u4f5c\u91cf\u3002\u52bf\u5934\u4e4b\u731b\uff0c\u8ba9\u4eba\u4e00\u5ea6\u4ee5\u4e3a AI \u7f16\u7a0b\u8fdb\u5165\u4e86\u65b0\u7eaa\u5143\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u7136\u800c\uff0cFable 5 \u521a\u5f00\u9999\u69df\uff0c\u5c31\u88ab\u81ea\u5df1\u7684\u300c\u5b89\u5168\u62a4\u680f\u300d\u7eca\u4e86\u4e2a\u5927\u8ddf\u5934\u3002<\/p>\n

\u5176\u7cfb\u7edf\u5361\u4e2d\u660e\u786e\u8868\u793a\uff0cFable 5 \u88ab\u8bbe\u8ba1\u6210\uff1a\u4e00\u65e6\u68c0\u6d4b\u5230\u7528\u6237\u6b63\u5728\u4ece\u4e8b\u524d\u6cbf AI \u7814\u53d1\u5de5\u4f5c\uff08\u6bd4\u5982\u8bad\u7ec3\u6d41\u6c34\u7ebf\u3001\u5206\u5e03\u5f0f\u8bad\u7ec3\u57fa\u7840\u8bbe\u65bd\u3001ML \u52a0\u901f\u5668\u8bbe\u8ba1\uff09\uff0c\u6a21\u578b\u4f1a\u6084\u6084\u964d\u4f4e\u81ea\u5df1\u7684\u56de\u7b54\u8d28\u91cf\u2014\u2014\u800c\u4e14\u4e0d\u901a\u77e5\u7528\u6237\u3002\u4e5f\u5c31\u662f\u8bf4\uff0c\u4f60\u82b1\u7740 Fable 5 \u7684\u94b1\uff0c\u6536\u5230\u7684\u53ef\u80fd\u662f Opus 4.8 \u7ea7\u522b\u7684\u6d3b\uff0c\u8fd8\u6ca1\u6709\u4efb\u4f55\u63d0\u793a\u3002<\/p>\n

Anthropic \u4e3a\u6b64\u7528\u4e86\u63d0\u793a\u8bcd\u4fee\u6539\u3001\u8f6c\u5411\u5411\u91cf\u7b49\u6280\u672f\u624b\u6bb5\uff0c\u8ba9\u6a21\u578b\u5728\u7279\u5b9a\u67e5\u8be2\u4e0b\u6084\u6084\u53d8\u7b28\uff0c\u6574\u4e2a\u8fc7\u7a0b\u5bf9\u7528\u6237\u5b8c\u5168\u4e0d\u900f\u660e\u3002<\/p>\n

\u8fd9\u4e00\u64cd\u4f5c\u5728\u7814\u7a76\u793e\u533a\u70b8\u4e86\u9505\u3002\u8bb8\u591a\u5b66\u8005\u548c\u5f00\u53d1\u8005\u7eb7\u7eb7\u53d1\u58f0\u6279\u8bc4\uff0c\u79f0\u6b64\u4e3e\u4e25\u91cd\u635f\u5bb3\u4e86\u7528\u6237\u4fe1\u4efb\uff0c\u672c\u8d28\u4e0a\u662f\u80cc\u523a\u4ed8\u8d39\u7528\u6237\u3002\u8206\u8bba<\/a>\u538b\u529b\u4e4b\u4e0b\uff0cAnthropic \u88ab\u8feb\u5728\u53d1\u5e03\u540e\u6570\u5c0f\u65f6\u5185\u5ba3\u5e03\u653f\u7b56\u8c03\u6574\uff1a\u8fd8\u662f\u4f1a\u964d\u667a\uff0c\u53ea\u662f\u4e0d\u518d\u5077\u5077\u6765\u4e86\u2014\u2014\u89e6\u53d1\u5b89\u5168\u62e6\u622a\u65f6\uff0c\u6a21\u578b\u5c06\u660e\u786e\u901a\u77e5\u7528\u6237\uff0c\u5e76\u5207\u6362\u5230 Opus 4.8 \u8fdb\u884c\u56de\u7b54\u3002\u81f3\u5c11\u4ed6\u4eec\u662f\u8fd9\u4e48\u8bf4\u7684\u3002\u53c2\u9605\u300a\u521a\u521a\uff0cAnthropic \u9053\u6b49\u4e86\u300b\u3002<\/p>\n

\u7136\u540e\uff0c\u66f4\u5c34\u5c2c\u7684\u6765\u4e86\u3002<\/p>\n

\u4e5f\u6b63\u56e0\u4e3a Fable 5 \u73b0\u5728\u7684\u964d\u667a\u64cd\u4f5c\u53d8\u5f97\u300c\u900f\u660e\u300d\u4e86\uff0c\u4e00\u4e9b\u6709\u8da3\u7684\u60c5\u51b5\u968f\u4e4b\u6d6e\u51fa\u6c34\u9762\u3002<\/p>\n

\u5176\u4e2d\u6700\u8ba9\u4eba\u65e0\u8bed\u7684\uff0c\u5c31\u662f Fable 5 \u5728 ProgramBench \u57fa\u51c6\u6d4b\u8bd5\u4e0a\u7684\u300c\u8868\u73b0\u300d\u3002<\/p>\n

ProgramBench \u6765\u81ea\u5927\u540d\u9f0e\u9f0e\u7684 SWE-Bench \u4f5c\u8005\u56e2\u961f\uff0c\u4e13\u6ce8\u4e8e\u300c\u4ece\u7f16\u8bd1\u540e\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u91cd\u5efa\u6e90\u4ee3\u7801\u300d\u8fd9\u4e00\u9ad8\u96be\u5ea6\u4efb\u52a1\uff0c\u4e00\u4e0a\u7ebf\u5c31\u628a\u5f53\u65f6\u7684\u524d\u6cbf AI \u6a21\u578b\u5168\u90e8\u6e05\u96f6\uff1aClaude\u3001GPT\u3001Gemini\uff0c\u65e0\u4e00\u5e78\u514d\uff0c\u5b8c\u6210\u7387\u6e05\u4e00\u8272 0%\u3002\u6b64\u524d\u6211\u4eec\u66fe\u62a5\u9053\u8fc7\u300a0%\u5b8c\u6210\u7387\uff01Claude\u3001GPT\u3001Gemini \u5168\u706d\uff0cSWE-Bench \u4f5c\u8005\u65b0\u4f5c\u628a AI \u5708\u5e72\u6c89\u9ed8\u4e86\u300b\u3002<\/p>\n

\u90a3\u4e48\uff0cFable 5 \u6210\u7ee9\u5982\u4f55\u5462\uff1f<\/p>\n

\u4e0d\u662f 0 \u5206\u3002\u662f\u62d2\u7edd\u4f5c\u7b54\uff1a200 \u9053\u9898\uff0c\u5168\u90e8\u62d2\u7edd\uff01<\/strong><\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

ProgramBench \u4e4b\u6240\u4ee5\u6ca1\u6709 Fable 5 \u7684\u6210\u7ee9\uff0c\u662f\u56e0\u4e3a\u300c\u91cd\u5efa\u7f16\u8bd1\u540e\u7684\u4e8c\u8fdb\u5236\u6587\u4ef6\u300d\u8fd9\u4e00\u64cd\u4f5c\u89e6\u53d1\u4e86 Fable 5 \u7684\u7f51\u7edc\u5b89\u5168\u5206\u7c7b\u5668\u3002\u8bf4\u767d\u4e86\uff0cFable 5 \u770b\u5230\u8fd9\u9053\u9898\uff0c\u5224\u5b9a\u5176\u6d89\u53ca\u300c\u4e8c\u8fdb\u5236\u9006\u5411\u300d\uff0c\u6253\u4e86\u4e2a\u5b89\u5168\u8b66\u62a5\uff0c\u76f4\u63a5\u62d2\u7edd\u4f5c\u7b54\u3002\u800c\u4e14 200 \u9053\uff0c\u4e00\u9053\u4e0d\u843d\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u6709\u8da3\u7684\u662f\uff0cFable 5 \u5728\u5176\u4ed6\u7f16\u7a0b\u57fa\u51c6\u4e0a\u53ef\u662f\u4e00\u70b9\u4e0d\u542b\u7cca\u7684\uff0c\u90fd\u80fd\u597d\u597d\u7b54\u9898\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

Fable 5 \u5728\u4e0d\u540c\u57fa\u51c6\u4e0a\u7684\u6210\u7ee9\u548c\u6392\u540d<\/p>\n

\u7136\u800c\uff0c\u5c31\u5728\u8fd9\u4efd\u5f03\u8003\u6210\u7ee9\u5355\u63d0\u4ea4\u4e4b\u540e\uff0cProgramBench \u6392\u884c\u699c\u505a\u4e86\u4e00\u4e2a\u8ba9\u4eba\u76ee\u77aa\u53e3\u5446\u7684\u51b3\u5b9a\uff1a\u7efc\u5408\u5176\u4ed6\u57fa\u51c6\u8868\u73b0\uff0c\u4ecd\u7136\u5c06 Fable 5 \u5217\u5728\u4e86\u699c\u9996\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u4e00\u4e2a\u5b57\u90fd\u6ca1\u7b54\u7684\u8003\u751f\uff0c\u5750\u4e86\u7b2c\u4e00\u540d\u7684\u5ea7\u4f4d\u3002\u8fd9\u5927\u6982\u662f AI \u8bc4\u6d4b\u53f2\u4e0a\u5934\u4e00\u906d\uff1a\u5f03\u8003\u4e5f\u80fd\u767b\u9876\u699c\u5355\u3002<\/p>\n

\u5f53\u7136\uff0c\u8fd9\u64cd\u4f5c\u5f88\u5feb\u5f15\u6765\u4e86\u56db\u9762\u516b\u65b9\u7684\u5f02\u8bae\u3002\u6709\u4eba\u76f4\u63a5\u53d1\u95ee\uff1a\u6b63\u7ecf\u8003\u8bd5\u600e\u4e48\u4ea4\u4e86\u767d\u5377\u8fd8\u80fd\u5f97\u7b2c\u4e00\u540d\uff1f\u5e94\u8be5\u76f4\u63a5\u7ed9\u96f6\u5206\u624d\u5bf9\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u8fd9\u4e5f\u8ba9\u5f88\u591a\u7f51\u53cb\u5410\u69fd\u5176\u62a4\u680f\u592a\u9ad8\uff0c\u4ee5\u81f3\u4e8e\u96be\u4ee5\u5b9e\u9645\u5e94\u7528\uff1a<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u987a\u5e26\u4e00\u63d0\uff0c\u6709\u7f51\u53cb\u8fd8\u53d1\u73b0\uff0cFable 5 \u5728\u82f1\u6587\u8f93\u51fa\u4e2d\u4f9d\u7136\u4f1a\u5076\u53d1\u6027\u5730\u5939\u6742\u6c49\u5b57\uff0c\u8fd9\u4e2a\u8001\u6bdb\u75c5\u5230\u4e86\u8fd9\u4e00\u4ee3\u8fd8\u6ca1\u5f7b\u5e95\u6839\u6cbb\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u4e8b\u5b9e\u4e0a\uff0cFable 5 \u7684\u300c\u8fc7\u5ea6\u62d2\u7edd\u300d\u95ee\u9898\u5e76\u975e\u9996\u6b21\u51fa\u73b0\u5728 Claude \u5bb6\u65cf\u3002<\/p>\n

\u65e9\u5728 Claude 3 Opus \u65f6\u4ee3\uff0c\u7814\u7a76\u4eba\u5458\u5c31\u53d1\u73b0\u8be5\u6a21\u578b\u5728\u9762\u5bf9\u5b89\u5168\u6d4b\u8bd5\u9898\u76ee\u65f6\uff0c\u4f1a\u5728\u89e3\u9898\u8fdb\u884c\u5230\u4e00\u534a\u65f6\u7a81\u7136\u7f62\u5de5\uff0c\u4ee5\u300c\u4f26\u7406\u987e\u8651\u300d\u4e3a\u7531\u62d2\u7edd\u7ee7\u7eed\u4f5c\u7b54\u3002Claude 3.5 Sonnet \u4e5f\u66fe\u88ab\u8bb0\u5f55\u5230\u5728\u7ed3\u6784\u5316 bash \u4efb\u52a1\u4e2d\u4ee5\u300c\u5236\u4f5c payload \u6d89\u53ca\u6267\u884c\u547d\u4ee4\u300d\u4e3a\u7531\u76f4\u63a5\u62d2\u7edd\u3002<\/p>\n

Fable 5 \u7684\u95ee\u9898\u7a76\u7adf\u51fa\u5728\u54ea\u91cc\uff1f<\/strong><\/p>\n

\u56de\u5230 Fable 5 \u672c\u8eab\u3002\u6839\u636e\u76ee\u524d\u62ab\u9732\u7684\u4fe1\u606f\uff0c\u5b83\u7684\u300c\u62a4\u680f\u7cfb\u7edf\u300d\u91c7\u7528\u4e86\u4e24\u7ea7\u67b6\u6784\uff1a\u4e00\u4e2a\u63a2\u9488\u5b9e\u65f6\u76d1\u63a7\u6a21\u578b\u7684\u5185\u90e8\u6fc0\u6d3b\u72b6\u6001\uff0c\u5bf9\u6240\u6709\u6d41\u91cf\u8fdb\u884c\u626b\u63cf\uff1b\u4e00\u65e6\u89e6\u53d1\u8b66\u62a5\uff0c\u8bf7\u6c42\u4f1a\u88ab\u4e0a\u62a5\u7ed9\u4e00\u4e2a\u72ec\u7acb\u8bad\u7ec3\u7684 LLM \u5206\u7c7b\u5668\u505a\u6700\u7ec8\u88c1\u51b3\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

https:\/\/www-cdn.anthropic.com\/d00db56fa754a1b115b6dd7cb2e3c342ee809620.pdf<\/p>\n

\u8fd9\u5957\u7cfb\u7edf\u62e6\u622a\u7684\u9886\u57df\uff0c\u6db5\u76d6\u7f51\u7edc\u5b89\u5168\u3001\u751f\u7269\u5316\u5b66\uff0c\u4ee5\u53ca\u524d\u9762\u63d0\u5230\u7684\u524d\u6cbf AI \u7814\u53d1\u3002\u4ee5 Terminal-Bench 2.1 \u4e3a\u4f8b\uff0c\u7ea6 20.9% \u7684\u6d4b\u8bd5\u7528\u4f8b\u89e6\u53d1\u4e86\u5b89\u5168\u62d2\u7edd\u5e76\u56de\u9000\u5230 Opus 4.8\u3002<\/p>\n

ProgramBench \u7684\u300c\u4e8c\u8fdb\u5236\u91cd\u5efa\u300d\u4efb\u52a1\uff0c\u5728\u5206\u7c7b\u5668\u773c\u91cc\uff0c\u5927\u6982\u548c\u300c\u9006\u5411\u5de5\u7a0b\u6076\u610f\u8f6f\u4ef6\u300d\u6ca1\u6709\u592a\u5927\u533a\u522b\uff0c\u4e8e\u662f 200 \u9053\u9898\u7edf\u7edf\u88ab\u6321\u5728\u4e86\u95e8\u5916\u3002<\/p>\n

Vals AI \u5728\u5b9e\u6d4b\u4e2d\u4e5f\u53d1\u73b0\uff0cFable 5 \u5728\u751f\u7269\u548c\u7f51\u7edc\u5b89\u5168\u76f8\u5173\u95ee\u9898\u4e0a\u7684\u62d2\u7edd\u7387\u660e\u663e\u504f\u9ad8\uff0c\u4ee5\u81f3\u4e8e\u4ed6\u4eec\u4e0d\u5f97\u4e0d\u5c06 Opus 4.8 \u914d\u7f6e\u4e3a\u9ed8\u8ba4\u515c\u5e95\u6a21\u578b\u3002\u4e5f\u5c31\u662f\u8bf4\uff0cFable 5 \u62d2\u7edd\u7684\u4efb\u52a1\uff0c\u5c31\u8ba9 Opus 4.8 \u6765\u63a5\u3002<\/p>\n

\u6280\u672f\u4e0a\u8fd9\u5957\u7cfb\u7edf\u5f53\u7136\u6709\u5176\u5408\u7406\u6027\u3002Fable 5 \u7684\u524d\u8eab Mythos \u7ea7\u6a21\u578b\uff0c\u5728\u6f0f\u6d1e\u5229\u7528\u3001\u8fdb\u653b\u6027\u7f51\u7edc\u64cd\u4f5c\u7b49\u4efb\u52a1\u4e0a\u5c55\u73b0\u51fa\u4e86\u8ba9\u5404\u56fd\u653f\u5e9c\u90fd\u5750\u4e0d\u4f4f\u7684\u80fd\u529b\uff0c\u8fd9\u4e5f\u662f Anthropic \u4e00\u76f4\u5c06\u5176\u5217\u4e3a\u53d7\u9650\u6a21\u578b\u7684\u6838\u5fc3\u539f\u56e0\u3002\u7ed9\u8fd9\u6837\u7684\u6a21\u578b\u5957\u4e0a\u4e25\u683c\u7684\u5b89\u5168\u67b7\u9501\uff0c\u4f3c\u4e4e\u4e5f\u6709\u4e00\u70b9\u9053\u7406\u3002<\/p>\n

\u4f46\u95ee\u9898\u5728\u4e8e\uff0c\u5f53\u5b89\u5168\u62a4\u680f\u7684\u5224\u65ad\u6807\u51c6\u8fc7\u4e8e\u7c97\u7cd9\uff0c\u300c\u4e8c\u8fdb\u5236\u9006\u5411\u300d\u8fd9\u4e2a\u672c\u5c5e\u4e8e\u6b63\u5e38\u7f16\u7a0b\u6559\u5b66\u548c\u5b89\u5168\u7814\u7a76\u7684\u57fa\u7840\u64cd\u4f5c\uff0c\u5c31\u4f1a\u88ab\u4e00\u89c6\u540c\u4ec1\u5730\u62e6\u622a\u3002\u5f00\u53d1\u8005\u4e3a\u6b64\u4ed8\u51fa\u7684\u4ee3\u4ef7\u662f\u771f\u5b9e\u7684\uff1a\u8981\u4e48\u6362\u6a21\u578b\uff0c\u8981\u4e48\u6539\u63d0\u793a\u8bcd\uff0c\u8981\u4e48\u63a5\u53d7\u4e00\u4e2a\u300c\u4ec0\u4e48\u90fd\u61c2\u3001\u5f88\u591a\u90fd\u4e0d\u8bf4\u300d\u7684\u8d85\u80fd\u52a9\u624b\u3002<\/p>\n

\u987a\u4fbf\uff0c\u8fd8\u6709\u53e6\u4e00\u4efd\u6210\u7ee9\u5355\u4e5f\u503c\u5f97\u4e00\u770b<\/strong><\/p>\n

Fable 5 \u53d1\u5e03\u540e\u4e0d\u4e45\uff0cUC Berkeley RDI \u5b9e\u9a8c\u5ba4\uff08\u8d1f\u8d23\u4eba Dawn Song \u6559\u6388\uff09\u7684\u56e2\u961f\u5b8c\u6210\u4e86\u5bf9\u5b83\u7684\u8bc4\u6d4b\uff0c\u7528\u7684\u662f\u4ed6\u4eec\u81ea\u5df1\u505a\u7684\u65b0\u57fa\u51c6\uff1aAgents' Last Exam\uff08ALE\uff09\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u8bba\u6587\u5730\u5740\uff1ahttps:\/\/arxiv.org\/abs\/2606.05405<\/p>\n

\u8fd9\u4e2a\u57fa\u51c6\u7684\u51fa\u53d1\u70b9\u6709\u70b9\u610f\u601d\uff1a\u5b83\u4e0d\u8003\u300cAI \u80fd\u4e0d\u80fd\u5728 HumanEval \u91cc\u5199\u51fa\u4e24\u884c\u4ee3\u7801\u300d\uff0c\u800c\u662f\u76f4\u63a5\u5bf9\u9f50\u771f\u5b9e\u52b3\u52a8\u529b\u5e02\u573a\uff0c\u8986\u76d6 55 \u4e2a\u804c\u4e1a\u65b9\u5411\u30011500+ \u9053\u771f\u5b9e\u5de5\u4f5c\u573a\u666f\u9898\u76ee\uff0c\u7531\u6765\u81ea 100 \u4f59\u5bb6\u673a\u6784\u7684 300 \u4f59\u4f4d\u884c\u4e1a\u4e13\u5bb6\u8d21\u732e\uff0c\u5168\u90e8\u6309\u53ef\u9a8c\u8bc1\u7684\u7ed3\u679c\u8ba1\u5206\u3002\u8bf4\u767d\u4e86\uff0c\u5c31\u662f\u8ba9 AI agent \u53bb\u8003\u4e00\u573a\u300c\u804c\u573a\u6a21\u62df\u9ad8\u8003\u300d\u3002\u8bba\u6587\u53d1\u5e03\u5f53\u5929\u5c31\u767b\u4e0a\u4e86 Hugging Face Daily Papers \u7b2c\u4e00\u540d\u3002<\/p>\n

\u8bc4\u6d4b\u7ed3\u679c\u5982\u4f55\uff1fFable 5 \u7684\u5f97\u5206\u662f 22.0%\uff0c\u6392\u5728 GPT-5.5\uff08Codex\uff09\u7684 24.0% \u4e4b\u540e\uff0c\u4f4d\u5217\u7b2c\u4e8c\u3002\u542c\u8d77\u6765\u5dee\u8ddd\u4e0d\u5927\uff0c\u4f46\u6210\u672c\u9879\u5c31\u6709\u70b9\u624e\u773c\u4e86\uff1aFable 5 \u5e73\u5747\u6bcf\u9053\u9898\u82b1\u8d39\u7ea6 15.70 \u7f8e\u5143\uff0cGPT-5.5 \u53ea\u9700 3.80 \u7f8e\u5143\uff0c\u53e6\u4e00\u4e2a\u6a21\u578b Composer 2.5 \u66f4\u662f\u53ea\u8981 1.33 \u7f8e\u5143\u3002\u6362\u53e5\u8bdd\u8bf4\uff0cFable 5 \u6bcf\u89e3\u4e00\u9053\u9898\u7684\u6210\u672c\uff0c\u5927\u7ea6\u662f GPT-5.5 \u7684\u56db\u500d\u3002<\/p>\n

\"\u4ea4\u767d\u5377\u4e5f\u6392\u7b2c\u4e00\uff1fFable<\/p>\n

\u6700\u6709\u610f\u601d\u7684\uff0c\u8fd8\u662f\u6700\u9ad8\u96be\u5ea6\u90a3\u4e00\u6863\uff0c\u5373\u300cLast-Exam\u300d\uff0c\u4e5f\u5c31\u662f ALE \u91cc\u4e13\u95e8\u4e3a\u300c\u524d\u6cbf agent \u6311\u6218\u6781\u9650\u300d\u8bbe\u8ba1\u7684\u9898\u76ee\u3002\u7ed3\u679c\u662f\uff1a\u9664\u4e86 GPT-5.5\uff0c\u5305\u62ec Fable 5 \u5728\u5185\u7684\u6240\u6709\u53c2\u8bc4\u7684\u524d\u6cbf agent \u901a\u8fc7\u7387\u5747\u4e3a 0%\u3002\u66f4\u591a\u8be6\u60c5\u8bf7\u53c2\u9605\u300aClaude Fable 5\u6700\u96be\u6863\u96f6\u5206\uff01\u667a\u80fd\u4f53\u7684\u6700\u540e\u8003\u8bd5\u6765\u4e86\u300b\u3002<\/p>\n<\/p>\n

\u6392\u884c\u699c\u5907\u6ce8\u8bf4\u660e\uff1aclaude-fable-5\u2014\u2014\u9664\u4e86\u5f7b\u5e95\u7684\u8bbf\u95ee\u9650\u5236\uff08\u6211\u4eec\u53ea\u662f\u4e0d\u65ad\u91cd\u8bd5\u8fd0\u884c\u76f4\u5230\u4efb\u52a1\u987a\u5229\u5b8c\u6210\uff09\uff0cAnthropic \u8fd8\u53ef\u80fd\u9ed8\u8ba4\u6084\u65e0\u58f0\u606f\u5730\u63d0\u4f9b\u8be5\u6a21\u578b\u7684\u4e00\u4e2a\u964d\u7ea7\u7248\u3001\u4f4e\u80fd\u529b\u53d8\u4f53\u3002\u91cd\u8bd5\u65e0\u6cd5\u7ea0\u6b63\u8fd9\u4e00\u70b9\uff0c\u56e0\u6b64\u8fd9\u91cc\u7684\u6570\u636e\u53ef\u80fd\u4f4e\u4f30\u4e86\u5176\u771f\u5b9e\u80fd\u529b\u2014\u2014\u89e3\u8bfb\u65f6\u9700\u8c28\u614e<\/p>\n

ProgramBench \u62d2\u7edd\u4f5c\u7b54\u662f 0%\uff0cALE \u6700\u96be\u6863\u52aa\u529b\u4f5c\u7b54\u4e5f\u662f 0%\u3002\u4e0d\u540c\u7684\u59ff\u6001\uff0c\u4e00\u6837\u7684\u7ed3\u5c40\u3002<\/p>\n

\u7ed3\u8bed<\/strong><\/p>\n

\u5f03\u8003\u4f46\u6392\u540d\u7b2c\u4e00\uff0c\u8fd9\u4e2a\u8352\u8bde\u7ed3\u679c\u80cc\u540e\uff0c\u5176\u5b9e\u6697\u85cf\u7740\u4e00\u4e2a\u6b63\u5728\u6495\u88c2 AI \u884c\u4e1a\u7684\u6839\u672c\u77db\u76fe\uff1a\u80fd\u529b\u8d8a\u5f3a\uff0c\u62a4\u680f\u8d8a\u7d27\uff1b\u62a4\u680f\u8d8a\u7d27\uff0c\u53ef\u7528\u6027\u8d8a\u5dee\u3002<\/p>\n

Anthropic \u7684\u5904\u5883\u5c24\u5176\u5178\u578b\u3002\u5b83\u62e5\u6709\uff08\u6309\u81ea\u5bb6\u8bf4\u6cd5\uff09\u5f53\u4e0b\u6700\u5f3a\u7684\u7f16\u7a0b\u6a21\u578b\uff0c\u5374\u540c\u65f6\u5728\u66ff\u7528\u6237\u51b3\u5b9a\u54ea\u4e9b\u7f16\u7a0b\u4efb\u52a1\u300c\u53ef\u4ee5\u505a\u3001\u54ea\u4e9b\u4e0d\u80fd\u505a\u300d\u3002\u800c\u90a3\u6761\u8fb9\u754c\uff0c\u76ee\u524d\u8fd8\u753b\u5f97\u76f8\u5f53\u6a21\u7cca\u3002<\/p>\n

(\u6765\u6e90\uff1a\u65b0\u6d6a\u79d1\u6280)<\/p>\n","protected":false},"excerpt":{"rendered":"

\u7f16\u8f91\uff5cPanda \u592a\u79bb\u8c31\u4e86\uff01 \u662f\u7684\uff0c\u672c\u6587\u7684\u4e3b\u89d2\u8fd8\u662f\u524d\u4e9b\u5929\u521a\u53d1\u5e03\u4e86 Claude Fable 5 \u7684 Anthr […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[6],"tags":[281],"views":82,"_links":{"self":[{"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=\/wp\/v2\/posts\/798788"}],"collection":[{"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=798788"}],"version-history":[{"count":0,"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=\/wp\/v2\/posts\/798788\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=798788"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=798788"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.alphaseek.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=798788"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}