ガロア第一論文と乗数イデアル他関連資料スレ18 (484レス)
上下前次1-新
291: 132人目の素数さん [] 07/02(水)16:33 ID:uMwR74wY(1/2)
>>290
集合族の共通部分は分かったのか?
クダラナイこと言ってないで勉強しろアホ
292(1): 132人目の素数さん [] 07/02(水)16:35 ID:uMwR74wY(2/2)
てか共通部分なんて数学の初歩の初歩の初歩
それすら分からんサルがなんで数学板に住み着いたのか
293: 132人目の素数さん [sage] 07/04(金)10:56 ID:5vz7yJnZ(1)
>>292
>(現代数学の系譜 雑談 ◆yH25M02vWFhPが)なんで数学板に住み着いたのか
大学1年の微分積分と線型代数が理解できなかったのがよほど悔しかったんでしょう、と邪推
294: 現代数学の系譜 雑談 ◆yH25M02vWFhP [] 07/05(土)12:39 ID:M5qP1slu(1/3)
これいいね
https://xenospectrum.com/what-is-pochomkin-understanding-the-decisive-weakness-of-llm-exposed-by-harvard-university-and-others/
XenoSpectrum
AIは「賢いフリ」をしていた──ハーバード大などが暴いたLLMの決定的弱点「ポチョムキン理解」とは?
Y Kobayashi
2025年7月4日
大規模言語モデル(LLM)は、流暢な会話をこなし、専門的な質問にも答える。その驚くべき能力に、私たちは「AIは本当に理解しているのではないか」という期待を抱きがちだ。しかし、その知性は本物なのだろうか? こうした我々の抱く最もな疑問に対し、2025年6月26日に発表された一つの研究論文が1つの答えを与えてくれるかもしれない。
ハーバード大学、MIT、シカゴ大学の世界トップクラスの研究者たちが共同で発表したこの論文は、LLMが概念を「理解しているフリ」をするという、根源的な欠陥を白日の下に晒した。本稿では、この「ポチョムキン理解」と名付けられた現象の核心に迫り、それが私たちのAIに対する見方、そして人工知能の未来に何を意味するのかを解説する。
賢いフリをするAI──「ポチョムキン理解」という新たな弱点
AIは「知っている」のに「できない」――韻律詩が暴いた矛盾
この一連のやり取りは、人間からすれば到底理解しがたい。
1.概念を正確に説明できる。
2.しかし、その概念を実践できない。
3.さらに、自分の実践が間違っていることを正しく認識できる。
これは人間には到底考えられない、深刻な「内部的な非一貫性(internal incoherence)」を示している。説明する知識(宣言的知識)と、それを使う知識(手続き的知識)が完全に分離してしまっている。まるで、料理のレシピを暗唱できるが、キッチンに立つと何も作れない料理人のようだ。しかも、自分が作った料理がレシピ通りでないことだけは正確に分かる、という奇妙な状況なのである。
主要LLMは軒並み不合格?驚きの実験結果
なぜベンチマークは「見せかけの理解」を見抜けなかったのか
AGIへの遠い道のり―業界に突きつけられた課題
私たちはAIの「知性」とどう向き合うべきか
私たちは、AIが生成する滑らかな言葉の裏にある「張りぼて」の可能性を常に意識し、その答えを鵜呑みにせず、批判的な視点を持って対話する必要がある。AIの「知性」の正体を見極め、その限界を理解した上で賢く付き合っていくこと。それこそが、これからの時代を生きる私たち一人ひとりに求められるリテラシーなのかもしれない。
295: 132人目の素数さん [] 07/05(土)14:54 ID:tHLA7S63(1)
AIじゃない生身の人間のコピペハリボテは速攻でバレバレだったけどな
296(1): 現代数学の系譜 雑談 ◆yH25M02vWFhP [] 07/05(土)18:07 ID:M5qP1slu(2/3)
虚数単位iと量子力学の話
これ面白い
https://www.nikkei-science.com/202508_078.html
日経サイエンス 2025年8月号
量子論に虚数は避けられないのか
M-O. ルヌー(仏国立情報学自動制御研究所サクレー研究センター) A. アシン(スペイン・光子科学研究所) M. ナバスクエス(オーストリア・量子光学・量子情報研究所)
3年前の2020年秋,著者の一人アシンは同じく著者の一人であるルヌーを,スペイン・バルセロナ近郊のカステルデフェルスにある光子科学研究所の自身の研究室に招いた。「あなたと議論したい問題がある」とアシンは切り出した。「ミゲル(ナバスクエス)と私が何年も取り組んできた問題だ」。興味津々の表情になったルヌーにアシンはいった。「標準的な量子論は虚数なしで成り立つだろうか?」
虚数は自身とかけ合わせると負になる数だ。哲学者デカルト(René Descartes)はそのような数を虚数と名づけ,彼が知っていて,その存在を受け入れていた,2乗しても負にならない数(現在,実数と呼ばれるもの)と区別した。その後,実数と虚数の和,すなわち複素数は,数学の複雑な問題を解く際に有用であることが認められ,数学者に広く受け入れられるようになった。しかしながら,複素数は物理学の基本理論の数式に不可欠な要素ではない。量子力学を除いては。
最も標準的な量子論の枠組みは複素数に依存している。この理論に現れる数を実数に限定すると新しい物理理論が得られる。「実数量子論」だ。21世紀の最初の10年間に,この実数版量子論が幅広い量子実験の結果を正しく記述できることがいくつかの研究チームによって示された。これらの発見により,多くの科学者が実数量子論によってあらゆる量子実験を説明できると信じるようになった。実数の代わりに複素数を使うという選択は物理的な立場を示しているのではなく,単に数学的な利便性の問題だと彼らは考えた。
だが,その予想は証明されていない。それが誤りである可能性はないのか? アシンの研究室でのやり取りの後,私たちは実数量子論を反証するための数カ月に及ぶ旅に出た。そして最終的に,実数量子論では説明できない結果を生じる量子実験を考案した。私たちの発見は,虚数が標準的な量子論の定式化において本質的な役割を果たしていることを意味している。つまり,虚数がなければ量子論は予測能力を失ってしまうのだ。これは何を意味するのか。虚数が何らかの形で実在することを示唆しているのだろうか。その答えは,標準的な量子論,ひいてはあらゆる物理理論を構成する要素が,実験結果を説明・予測するための単なる数学的道具ではなく「実在している」という考え方を,どの程度真摯に受け止めるかによる。
続きは2025年8月号の誌面でどうぞ。
著者
Marc-Olivier Renou / Antonio Acín / Miguel Navascués
ルヌーは仏国立情報学自動制御研究所(INRIA)サクレー研究センターに所属する理論物理学者。アシンはスペイン・光子科学研究所量子情報理論グループのリーダー。ナバスクエスはウィーンにある量子光学・量子情報研究所の若手グループのリーダー。
つづく
297(1): 132人目の素数さん [] 07/05(土)18:07 ID:M5qP1slu(3/3)
つづき
関連記事
「最終決着『ベルの不等式』の破れの実験」,R. ハンソン/ K. シャルム,日経サイエンス2019年2月号。
原題名
Imaginary Universe(SCIENTIFIC AMERICAN April 2023)
https://www.nikkei-science.com/?p=76146
英語で読む日経サイエンス
SCIENTIFIC AMERICAN April 2023
Imaginary Universe
量子論に虚数は避けられないのか
By Marc-Olivier Renou /Antonio Acín/Miguel Navascués M-O. ルヌー /A. アシン/M. ナバスクエス
https://www.scientificamerican.com/article/quantum-physics-falls-apart-without-imaginary-numbers/
April 1, 2023
13 min read
Quantum Physics Falls Apart without Imaginary Numbers
Imaginary numbers—the square roots of negative numbers—are an inescapable part of quantum theory, a study shows
By Marc-Olivier Renou, Antonio Acín & Miguel Navascués
(引用終り)
以上
298(1): 132人目の素数さん [] 07/06(日)06:43 ID:a2BajG8e(1/3)
武漢では開性定理の応用について
サーベイがあった
299(2): 132人目の素数さん [] 07/06(日)08:08 ID:+k1m9OFg(1/9)
>>298
>武漢では開性定理の応用について
>サーベイがあった
"開性定理"は、初耳です
google検索のAI回答は下記ですが、合ってますか?
(日本語だと情報が少ないが、英語などでは情報があるかもです)
google検索
多変数関数論 開性定理とは
<AI による概要>(AI の回答には間違いが含まれている場合があります)
多変数関数論における開性定理とは、多変数複素関数が局所的に正則であれば、その関数は開集合上で正則であるという定理です。より具体的に言うと、ある開集合内の点において、その点の近傍で正則な多変数複素関数は、その開集合全体で正則になるという定理です。
開性定理の概要:
多変数複素関数:複数の複素変数を入力とする関数です。例えば、f(z1, z2) のように表されます。
正則性:複素関数が微分可能であること、つまり、複素微分が定義できることを意味します。
局所正則性:ある点の近傍で正則であることです。
開集合:任意の点に対して、その点を含む十分小さな開円盤(近傍)がその集合に完全に含まれるような集合です。
大域正則性:開集合全体で正則であることです。
開性定理の重要性:
開性定理は、多変数複素関数論において、局所的な性質が、より広い範囲の集合における性質にどのように影響するかを示す重要な定理です。この定理によって、局所的に正則な関数を調べることで、その関数全体の性質を把握できる可能性が開かれます。
例:
例えば、ある多変数複素関数が、ある開集合Dのすべての点zで、その点の近傍で正則だとします。この時、開性定理により、その関数は開集合D全体で正則であることが保証されます。
関連する概念:
正則関数:複素微分可能な関数。
多変数複素関数論:複数の複素変数を扱う関数論。
岡潔:多変数複素関数論に大きな貢献をした数学者。
補足:
開性定理は、多変数複素関数論における基本的な定理の一つであり、様々な応用を持っています。例えば、複素解析の手法は、応用数学や物理学、工学など、幅広い分野で利用されています。
300(1): 132人目の素数さん [] 07/06(日)08:19 ID:+k1m9OFg(2/9)
>>299 補足
前記の
”日経サイエンス 2025年8月号
量子論に虚数は避けられないのか”>>296-297
これで、『実数量子論』(実2変数 (x,y)の関数論を使う)のと
1変数複素関数論 z=x+iy
との違いは、"開性定理"にありってことですかね? (^^
301(2): 132人目の素数さん [] 07/06(日)09:16 ID:a2BajG8e(2/3)
openness conjectureが解決された結果
開性定理が生まれた
そのeffective versionsが
複素幾何に応用されている
302: 132人目の素数さん [sage] 07/06(日)14:45 ID:esX0ReXc(1)
チンチョンチャイナマン
303: 132人目の素数さん [] 07/06(日)15:43 ID:/E5gvvZ5(1/5)
>>300
>『実数量子論』(実2変数 (x,y)の関数論を使う)のと
>1変数複素関数論 z=x+iyとの違いは、
>"開性定理"にありってことですかね?
頭わるそう
304(1): 132人目の素数さん [] 07/06(日)17:49 ID:+k1m9OFg(3/9)
>>301
google検索
Several Complex Variables openness conjecture openness theorem
検索結果(要約メモ:openness conjecture=Demailly's Strong Openness Conjecture らしい)
1)
https://english.cas.cn/bcas/2020_3/202102/P020210210712420301203.pdf
2 ページ
BCAS vol.34 No.3 2020 (Bulletin of the Chinese Academy of Sciences)
TKK Young Scientist Award in Mathematics and Physics
Solutions of Demailly’s Strong Openness Conjecture and Related Problems in Several Complex Variables
By YAN Fusheng (Staff Reporter)
The 2020 TKK Young Scientist Award in Mathematics & Physics went to Prof. GUAN Qi’an, a young talented mathematician from Peking University, for his solutions (joint with ZHOU Xiangyu) of a series of problems in several complex variables, particularly for his proof (joint with ZHOU Xiangyu) of the Demailly’s strong openness conjecture.
GUAN has mainly engaged himself in the study of several complex variables, which explores the properties and structures of holomorphic functions of several variables, and is also called complex analysis of several variables. Because the properties of holomorphic functions are largely affected by the geometric and topological properties of their domains of definition,the research involves not only the studying of local properties, but also of global properties.
In the research of several complex variables, various methods from partial differential equations, algebraic
geometry, complex geometry, topology, Lie groups and other areas are widely applied. The research of function theory of several complex variables has also driven the development of the above-mentioned research fields. For example, LU Qikeng proved “LU Qikeng Theorem” named after him; SIU Yum-Tong proved the deformational invariance of plurigenera of projective algebraic manifolds; ZHOU Xiangyu proved the extended future tube conjecture which was listed as an
unsolved problem in the Encyclopedia of Mathematics.
In cooperation with Prof. ZHOU Xiangyu, a CAS Member at the Academy of Mathematics and Systems Science (AMSS) under the Chinese Academy of Sciences,GUAN solved the optimal L2 extension problem,
proposing new ideas and methods. They established the optimal L2 extension theorem by unifying the previous
以下略
注)TKK:the Tan Kah Kee Young Scientist Award 陈嘉庚科学奖基金会 https://tsaf.cas.cn/en/
つづく
305(1): 132人目の素数さん [] 07/06(日)17:49 ID:+k1m9OFg(4/9)
つづき
2)
https://annals.math.princeton.edu/wp-content/uploads/annals-v182-n2-p05-p.pdf
A proof of Demailly's strong openness conjecture
Annals of Mathematics
2015/01/14 — In the present article, we discuss a more general conjecture — the strong openness conjecture about multiplier ideal sheaves for ...
12 ページ
3)
https://arxiv.org/pdf/2109.00353
arXiv:2109.00353v1 [math.CV] 1 Sep 2021
Q Guan 著 · 2021 · 被引用数: 10 — The strong openness property is an important feature of multiplier ideal sheaves and used in the study of several complex variables, algebraic ...
4)
https://arxiv.org/pdf/2203.01648
arXiv:2203.01648v4 [math.CV] 1 Apr 2024
S Bao 著 · 2022 · 被引用数: 13 — of several complex variables, complex algebraic geometry and complex differential geometry (see e.g. [48, 42, 44, 12, 13, 11, 14, 39, 40, 45 ...
BOUNDARY POINTS, MINIMAL L2 INTEGRALS AND CONCAVITY PROPERTY
SHIJIE BAO, QI’AN GUAN, AND ZHENG YUAN
Abstract. For the purpose of proving the strong openness conjecture of
multiplier ideal sheaves, Jonsson-Mustat¸˘a posed an enhanced conjecture and
proved the two-dimensional case, which says that: the Lebesgue measure of
the set {cFo(ψ)ψ − log |F| < log r} divided by r2 has a uniform positive lower
bound independent of r, for a plurisubharmonic function ψ and a holomorphic
function F near the origin o. Jonsson-Mustat¸˘a’s conjecture was proved by
Guan-Zhou depending on the truth of the strong openness conjecture. However, it is still a question whether one can prove Jonsson-Mustat¸˘a’s conjecture
without using the strong openness property, and obtain a sharp effectiveness
result for this conjecture.
In this article, we use an L2 method with the weight functions ψ − log |F|
and firstly consider a module at at a boundary point of the sublevel sets of
a plurisubharmonic function. By studying the minimal L2
integrals on the sublevel sets of a plurisubharmonic function with respect to the module at the
boundary point, we establish a concavity property of the minimal L2 integrals.
As applications, we obtain a sharp effectiveness result related to JonssonMustat¸˘a’s conjecture, which completes the approach from the conjecture to the strong openness property. We also obtain a strong openness property of the module and a lower semi-continuity property with respect to the module.
つづく
306(1): 132人目の素数さん [] 07/06(日)17:50 ID:+k1m9OFg(5/9)
つづき
5)
https://www2.math.kyushu-u.ac.jp/~kusakabe/YMWSCV2024/abstracts.pdf
九大数理学研究院
The proof is based on a density formula of Zelditch, the Abel–Jacobi theory, Fekete points theory, and a new perturbation method. This is joint work (arXiv: ...
10 ページ
<関連箇所>
Young Mathematicians Workshop on Several Complex Variables 2024
Osaka Metropolitan University∗
13:40–14:20 Wang Xu (Sun Yat-sen University)
Optimal L2 extension of openness type and related topics
Abstracts
L2 extension theorems and optimal L2
extension theorems are important and powerful tools in several complex variables and complex geometry. There is a closely
related problem called the optimal L2
extension problem of openness type: given
a holomorphic section f defined on a neighbourhood U of a subvariety S, find a
holomorphic extension of f|S to the ambient manifold, whose L2 norm is optimally
controlled by the L2 norm of f on U. In this talk, I will present a solution on
weakly pseudoconvex K¨ahler manifolds, which generalizes a couple of known results. I will also discuss some connections and applications to related topics, such as sharper L2
extensions and generalized Suita conjectures. This talk is mainly
based on joint work with Prof. Xiangyu Zhou.
6)動画がある
https://www.mathnet.ru/eng/present37745
International Conference Dedicated to the 100th Anniversary of the Birthday of V. S. Vladimirov (Vladimirov-100)
January 10, 2023 14:00–14:30, Moscow, Steklov Mathematical Institute, room 430 (Gubkina 8) + Zoom
X. Zhou
Academy of Mathematics and Systems Science, Chinese Academy of Sciences
<動画アドレス>
https://www.mathnet.ru/PresentFiles/37745/37745.mp4
Abstract: We'll talk about some recent results in several complex variables and complex geometry, e.g., the solution of Demailly's strong openness conjecture on multiplier ideal sheaves, converse of L^2 existence theorem due to Hoermander and Demailly, and their applications in some fundamental problems in complex geometry.
7)余録(openness conjectureは出てこないが、それ以前の整理には役立つだろう)
https://www.ms.u-tokyo.ac.jp/~noguchi/Nog_remark_SCV_Kanazawa2024.pdf
Some remarks on basic materials in several complex variables
J. Noguchi
30th Symposium of Complex Geometry, Kanazawa 2024
(引用終り)
以上
307: とおりすがり [] 07/06(日)17:58 ID:bD6lgxLz(1)
実数論でも同値関係の概念が理解不能な
なトンデモ>1や量子力学と場の量子論の
違いが理解不能なもみ消し徘徊元教授に
は、量子力学の基礎は理解不能。
オカルトIUTがお似合いだ
308(1): 現代数学の系譜 雑談 ◆yH25M02vWFhP [] 07/06(日)18:05 ID:+k1m9OFg(6/9)
Copilot:
What is Demailly's Strong Openness Conjecture
回答:
Demailly's Strong Openness Conjecture is a deep result in complex geometry and analysis, particularly in the theory of multiplier ideal sheaves associated with plurisubharmonic functions.
The Conjecture (Now a Theorem)
Let \varphi be a plurisubharmonic function on a complex manifold X. The conjecture asserts that:
\mathcal{I}_+(\varphi) = \mathcal{I}(\varphi)
where:
- \mathcal{I}(\varphi) is the multiplier ideal sheaf: the sheaf of holomorphic functions f such that |f|^2 e^{-\varphi} is locally integrable.
- \mathcal{I}_+(\varphi) := \bigcup_{\varepsilon > 0} \mathcal{I}((1+\varepsilon)\varphi)
This means that the ideal sheaf stabilizes under small perturbations of the weight function \varphi, reflecting a kind of openness in the space of integrable functions
Historical Context
Jean-Pierre Demailly proposed this conjecture as a strengthening of the earlier Openness Conjecture (by Demailly and Kollár), which was concerned with the integrability of exponentials of plurisubharmonic functions.
✅ Resolution
The conjecture was proven by Qi'an Guan and Xiangyu Zhou in 2015. Their work not only settled the conjecture but also led to further insights into:
- The behavior of singularities of plurisubharmonic functions
- The structure of multiplier ideal sheaves
- Applications to algebraic geometry and complex differential geometry
You can explore the original proof in the Annals of Mathematics article1) or the expanded version on arXiv2).
Would you like to dive into the techniques used in the proof, such as the L^2 extension theorem or curve selection lemma? I think you'd find the analytic machinery quite elegant.
注:
1)Annals of Mathematics article
https://annals.math.princeton.edu/2015/182-2/p05
A proof of Demailly’s strong openness conjecture
Pages 605-616 from Volume 182 (2015), Issue 2 by Qi'an Guan, Xiangyu Zhou
Abstract
In this article, we solve the strong openness conjecture on the multiplier ideal sheaf associated to any plurisubharmonic function, which was posed by Demailly.
2)arXiv
https://arxiv.org/abs/1401.7158
[Submitted on 28 Jan 2014]
Strong openness conjecture and related problems for plurisubharmonic functions
Qi'an Guan, Xiangyu Zhou
309(2): 132人目の素数さん [] 07/06(日)18:05 ID:/E5gvvZ5(2/5)
素人のコピペより、AIの解説のほうがマシ(笑)
多変数複素関数論における openness theorem(開性定理) は、
複素多様体や複素解析空間の文脈で重要な結果で、
特に 正則関数の像に関する性質 を述べるものです。
以下にその概要を簡潔に説明します。
開性定理とは
開性定理は、多変数の正則関数(holomorphic function)の像が「開集合」になるという性質を保証する定理です。
具体的には、次のように述べられます:
定理(開性定理):
複素数空間 C^n の開集合 U⊂C^n から C^m への正則関数 f:U→C^mが与えられたとき、
もし f が 局所的に単射(locally injective)であるか、
またはヤコビ行列のランクがある点で最大である場合、
f の像 f(U) は C^m において開集合となる。
310(1): 132人目の素数さん [] 07/06(日)18:06 ID:/E5gvvZ5(3/5)
ポイント
局所的な単射性:f が局所的に単射であるとは、U の各点 p においてある近傍が存在し、その近傍内で f が単射(1対1)であることを意味します。
これは、ヤコビ行列 Df(p) が全単射(つまり、ランクが m)であることと密接に関係します。
開集合:像 f(U) が開集合であるとは、f(U) の各点の周りに C^mの開近傍が含まれることを意味します。
これは、直感的には f が「連続的に広がった像」を作ることを示しています。
多変数の特徴:1変数の場合、正則関数の開写像定理(open mapping theorem)はよく知られており、非定数正則関数は必ず開集合に写すことが保証されます。
多変数の場合は、局所的な単射性やヤコビ行列の条件が必要になります。これは、多変数の正則関数の振る舞いが1変数の場合よりも複雑だからです。
上下前次1-新書関写板覧索設栞歴
あと 174 レスあります
スレ情報 赤レス抽出 画像レス抽出 歴の未読スレ AAサムネイル
ぬこの手 ぬこTOP 0.017s