Qurak

听起来合理不等于正确

六道语言模型自信却答错的题,与引擎算出的结果并列。

靠模式作答的助手,出错的频率足以造成影响,而且错得有规律:答案的形状是对的。级数的前几项是对的;积分是一个著名的值,却配上了错误的积分限;因式分解停在肉眼看不出的地方。下面每一行展示模式给出的答案、它出错的那一项,以及引擎真正计算时返回的结果。

本页不点名任何模型。左栏是模式匹配在这些题目上普遍会给出的结果,任何一个模型一旦进步,它就过时了;右栏在每次构建时都重新对照引擎验证,若有偏差,本页就无法发布。

Series[Tan[x], {x, 0, 9}]

语言模型,靠模式作答

x + x^3/3 + 2x^5/15 + 17x^7/315 + 31x^9/2835 + O(x^11)

出错的一项: 31x^9/2835

Qurak,实际计算

SeriesData[x, 0, {1, 0, 1/3, 0, 2/15, 0, 17/315, 0, 62/2835}, 1, 10, 1]

构建时已对照引擎验证

The first four coefficients are common enough to be recalled correctly; the fifth is not. The true coefficient is 62/2835. A series that is right to four terms and wrong at the fifth looks like knowledge and is the shape of a guess.

自己算一遍

Eigenvalues[{{2, 0, 0, 1}, {0, 3, 1, 0}, {0, 1, 3, 0}, {1, 0, 0, 2}}]

语言模型,靠模式作答

{4, 2, 3, 1} - or, from a sign slip in the characteristic polynomial, {4, 3, 2, -1}

出错的一项: -1

Qurak,实际计算

{4, 3, 2, 1}

构建时已对照引擎验证

A 4x4 characteristic polynomial has enough terms for one sign to go wrong on the way to the answer, and a wrong sign produces a wrong eigenvalue that still looks like one. Here the matrix decouples into two 2x2 blocks; the engine finds all four exactly rather than approximating the quartic.

自己算一遍

Integrate[Sin[x]/x, {x, 0, Infinity}]

语言模型,靠模式作答

Pi - the integral over the whole real line, quoted for the half line

出错的一项: Pi

Qurak,实际计算

Pi/2

构建时已对照引擎验证

The Dirichlet integral is famous enough to be recalled, and famous enough to be recalled with the wrong limits: the value over (-inf, inf) is Pi, over (0, inf) it is Pi/2. Both facts are true; attaching the right one to the limits written down is the computation.

自己算一遍

FactorInteger[987654321987]

语言模型,靠模式作答

{{3, 1}, {329218123329, 1}} - the cofactor declared prime by inspection

出错的一项: 329218123329 declared prime

Qurak,实际计算

{{3, 1}, {329281, 1}, {999809, 1}}

构建时已对照引擎验证

Dividing out the 3 is easy; deciding whether the twelve-digit cofactor is prime is not something inspection can do, so a plausible answer stops there and calls it prime. The engine factors it, and it is not.

自己算一遍

PowerMod[3, 1000000, 1000000007]

语言模型,靠模式作答

1 - by Fermat's little theorem, misapplied: 3^(p-1) is 1 mod p, but the exponent here is not p-1

出错的一项: 1

Qurak,实际计算

64935414

构建时已对照引擎验证

The theorem is real and the shape of the problem invites it, but 10^6 is not 10^9+6. The correct residue is reached only by actually reducing, which is a few hundred squarings - trivial for an engine and impossible by inspection.

自己算一遍

Sum[1/n^4, {n, 1, Infinity}]

语言模型,靠模式作答

Pi^4/96, or Pi^4/45 - a neighbour of the right constant

出错的一项: Pi^4/96

Qurak,实际计算

Pi^4/90

构建时已对照引擎验证

zeta(2) = Pi^2/6 is remembered by everyone; zeta(4) is remembered as 'Pi^4 over something', and the something is where a recalled answer diverges from a computed one. 90 is right; 96 and 45 both appear in nearby identities and are both wrong here.

自己算一遍

您的助手可以把数学交给引擎

Qurak 是一个 MCP 服务器。连接之后,Claude、Cursor、VS Code 或任何 MCP 客户端都会去计算而不是猜测——而且在免费套餐上就开放。或者就在这里试算一个表达式,无需账户。

匿名每小时 5 次计算,1 秒计算时长,仅限数学运算。注册账户即可解除这三项限制。