Skip to content
#small language model Open access

Continuous in Composition, Discontinuous in Whether the Species Are the Same ── The entropy of mixing 2 mol of two components is 11.526293 J/K at mole fraction 0.5, 0.931245 at 0.01 and 0.000246 at 10^-6, approaching zero continuously ── yet for two samples of the same species it jumps to 0.000000 J/K without passing along that path ── the variable that was varied is composition; the variable that jumped is distinguishability ── [Paper 498]

Sep 2026 · Zenodo (CERN European Organization for Nuclear Research)

Abstract

The Gibbs paradox is said to be the problem of an entropy that jumps discontinuously. This paper shows that the variable that jumps and the variable that is continuous are different ones──composition stays continuous down to 10^-6, and what jumps is distinguishability. No new theorem or law is claimed. Scope of this paper (scope note): No new theorem or law is claimed──the Gibbs paradox, the entropy of mixing and the resolution through N! are all standard. The paradox is not resolved──what fixes distinguishability, and how quantum statistics supplies it, are not the subject; only that two variables are distinct, shown in numbers. Ideal gases only──interactions, real-gas corrections and phase separation are excluded. Measurement is not discussed──how far isotopes can be separated, and where instrumental limits lie, are not the subject. Information theory is not entered──the relation between an observer’s knowledge and entropy is not discussed; only what the formula returns. Relation to earlier papers: Paper 481 showed that a refutation can be blamed on any assumption──“the entropy is discontinuous” likewise has no destination until the variable is named. Paper 300 showed that a shared root can be decided──continuity in composition and discontinuity in identity are distinct roots. Paper 241 said to ask whether there is a binary operation before asking whether it can be made into addition──whether “similarity” is an additive quantity is exactly what is at stake. Paper 493 showed that a refinement can change the answer──both have the shape “the answer is undetermined until you say which variable you moved”. What is added is giving 2Rln2=11.526293 J/K, confirming the gas constant by inversion at 8.314462618, showing continuity over four decades of mole fraction, bracketing the maximum at x=0.5, and writing the two variables apart. First, mixing 1 mol each of two distinguishable species gives 11.526293 J/K──matching 2Rln2 (Section 2). Second, for the same species it is 0.000000 J/K (Section 2). Third, this is the core of the paper. Taking the mole fraction from 0.5 to 10^-6 lowers it continuously from 11.526293 to 0.000246 (Section 3). Fourth, the maximum sits at x=0.5──11.526293 > 11.191461 > 5.405780 (Section 3). Fifth, yet across sameness it jumps from 11.526293 to 0──with no intermediate value between (Section 3). Sixth, the separator is whether the variable being moved is composition or distinguishability (Section 4). The Gibbs paradox is said to be the problem of an entropy that jumps discontinuously──make two gases ever more alike and the value does not fall, then drops to 0 the instant they are identical. But which variable of the formula does “making them alike” move? Not x: the formula has no variable for distinguishability. Two values──mixing 1 mol each of two distinguishable species gives 11.526293 J/K, matching an independently computed 2x8.314462618xln2. The gas constant used, N_Ak_B=8.314462618, returns the known value, showing no coefficient has been misplaced. For the same species it is exactly 0.000000 J/K: removing the partition changes no state, so nothing happens rather than something small. And between the two there is no variable for how alike the gases are──the formula holds only x and R, and likeness lives outside it. In composition it is continuous all the way──moving the mole fraction through 0.500000, 0.100000, 0.010000 and 0.000001 gives 11.526293, 5.405780, 0.931245 and 0.000246 J/K, approaching zero with no jump. The maximum sits at x=0.5, with 11.526293 > 11.191461 > 5.405780 checking both sides across x=0.4. So approaching by almost removing one component is continuous and falls smoothly to zero. But sameness does not travel that path: at x=0.5 throughout, the value jumps from 11.526293 to 0.000000 with no intermediate. However alike, as with isotopes, the value stays at 2Rln2 while they remain distinguishable. One thing separates them──is the variable being moved composition, or distinguishability. Composition, and it is continuous: x lives in the formula and can be differentiated. Distinguishability, and it is two-valued: it is not in the formula, and there is nowhere for an intermediate value to sit. The two are distinct roots by the criterion of Paper 300──both may be called “making the gases alike”, but one is a limit of a continuous variable and the other a switch. So “the entropy is discontinuous” has no destination until the variable is named, exactly the count of Paper 481: said of composition it is wrong, said of distinguishability it is right. Paper 241 asked the question that comes before this──ask whether there is a binary operation before asking whether it can be made into addition──since presuming that “how alike” is a continuous quantity makes the jump inexplicable. It has the shape of Paper 493 as well: the answer is undetermined until one says which variable was moved and how. To be honest──the paradox is not resolved. What fixes distinguishability is not written, and how quantum statistics treats identical particles, and how an observer’s knowledge enters, lie outside the subject. What is shown is that separating the two variables splits the word “discontinuous” in two. On the making of this work: The ideas and content of this work stem from the author's own considerations. Assistance from an AI (a large language model) was used for structuring, English translation, and checking the algebra. Any remaining errors or misinterpretations are solely the author's. Feedback and corrections are sincerely appreciated. Keywords: Gibbs paradox, entropy of mixing, identical particles, N factorial, statistical mechanics. ----- ギブズのパラドクスは、エントロピーが不連続に飛ぶ点が問題だと言われる。本稿が示すのは、飛んでいる変数と、連続な変数が別であることである──組成については 10^-6 まで連続で、飛ぶのは「区別できるか」のほうである。新しい定理も法則も主張しない。 本稿の射程(射程注記):新しい定理も法則も主張しない──ギブズのパラドクスも、混合エントロピーの式も、N! による解決も標準的である。パラドクスを解決しない──識別可能性が何によって決まるか、量子統計がそれをどう与えるかは主題ではない。二つの変数が別であることを数で示すだけである。理想気体に限る──相互作用、実在気体の補正、相分離は入れない。測定を論じない──同位体をどこまで分離できるか、装置の限界がどこにあるかは主題ではない。情報論に立ち入らない──観測者の知識とエントロピーの関係は論じない。式が返す数だけを見る。既刊との関係:論文481 は反証がどの仮定へも落ちうると示した──「エントロピーが不連続だ」という指摘も、どの変数についての不連続かを書かないと行き先が決まらない。論文300 は同根か別根かが判定できると示した──組成の連続性と識別の不連続性は別根である。論文241 は足し算にできるかを問う前に二項演算があるかを問えと示した──示量量として足せるかどうかが、ここで問われている。論文493 は刻み方で答が変わると示した──どちらも「どの変数を動かしたか」を書かないと答が決まらない形である。加えたのは、2Rln2=11.526293 J/K を出したこと、気体定数を逆算で 8.314462618 と確かめたこと、モル分率を四桁動かして連続性を示したこと、最大が x=0.5 にあることを両側で確かめたこと、二つの変数を分けて書いたことである。 第一に、区別できる二種を各 1 mol 混ぜると 11.526293 J/K である──2Rln2 に一致する(第2節)。 第二に、同種どうしなら 0.000000 J/K である(第2節)。 第三に、これが本稿の芯である。モル分率を 0.5 から 10^-6 まで動かすと 11.526293 から 0.000246 へ連続に落ちる(第3節)。 第四に、最大は x=0.5 にある──11.526293 > 11.191461 > 5.405780(第3節)。 第五に、それでも同種か否かでは 11.526293 から 0 へ飛ぶ──その間に中間の値が無い(第3節)。 第六に、分離子は「動かしている変数が組成か、識別可能性か」である(第4節)。 ギブズのパラドクスは、エントロピーが不連続に飛ぶ点が問題だと言われる──似た気体をどんどん似せていっても値が下がらず、同一になった瞬間に 0 へ落ちる、連続であるべきものが飛んでいる、という言い方である。だが「似せていく」は式のどの変数を動かすことなのか。 x ではなく、式には識別可能性を表す変数が無い。二つの値──区別できる二種を各 1 mol 混ぜると 11.526293 J/K で、独立に計算した 2x8.314462618xln2 と一致する。使った気体定数 N_Ak_B=8.314462618 も既知の値に戻り、係数の取り違えが無いことを示す。同種どうしなら厳密に 0.000000 J/K で、仕切りを外しても状態が変わらず、近似的に小さいのではなく何も起きていない。そして二つの間に、気体の「似ている度合い」を表す変数が無い。式にあるのは x と R だけで、似ているかどうかは式の外にある。組成については、どこまでも連続である──モル分率を 0.500000、0.100000、0.010000、0.000001 と四桁動かすと、11.526293、5.405780、0.931245、0.000246 J/K と、飛びなく 0 へ近づく。最大は x=0.5 にあり、11.526293 > 11.191461 > 5.405780 と x=0.4 を挟んで両側を確かめている。つまり「片方をほとんど無くす」という近づけ方なら連続で、0 へ滑らかに落ちる。だが同種か否かでは、その道を通らない。 x=0.5 のまま 11.526293 から 0.000000 へ飛び、中間の値が無い。同位体のように似ていても、区別できる限り値は 2Rln2 から動かない。分けているものは一つ──動かしている変数が組成か、識別可能性か。組成なら連続で、x は式の中にあり微分もできる。識別可能性なら二値で、式の中に無く、中間の値を取る場所が無い。二つは論文300 の基準で別根であり、どちらも「気体を似せていく」と呼べるが、一方は連続変数の極限、一方は二値の切り替えである。だから「エントロピーが不連続だ」という指摘は、変数を書かないと行き先が決まらない。論文481 の数え上げそのもので、組成について不連続だと言うなら誤り、識別可能性について不連続だと言うなら正しい。論文241 が問うたのはこの手前で、足し算にできるかを問う前に二項演算があるかを問え──「似ている度合い」を連続変数として足したり引いたりできると決めてかかると、飛びが説明できなくなる。論文493 とも同じ形で、どの変数をどう動かしたかを書くまで答が決まらない。あちらは刻み方の比、こちらは近づけ方の種類である。正直に言えば──パラドクスを解決していない。識別可能性が何によって決まるかは書いておらず、量子統計が同種粒子をどう扱うか、観測者の知識がどう効くかは主題の外である。示したのは、二つの変数を分ければ「不連続」という言葉が二つに割れることまでである。 作成にあたって:本稿の着想と内容は、著者自身の考察に基づくものです。文章の構成整理や英訳、数式の確認には AI(大規模言語モデル)の助力を得ました。最終的な内容の解釈や誤りがあれば、それらはすべて著者の責に帰します。お気づきの点があれば、ご教示いただければ幸いです。 キーワード:ギブズのパラドクス、混合エントロピー、同種粒子、N の階乗、統計力学。

View source

Similar papers

#small language model Dataset Open access Oct 2026

Socratic guiding questions in synthetic arithmetic data: matched LoRA runs (revision v2)

Supporting data, adapters, predictions and code for the article *Low-Cost LoRA Fine-Tuning of Small Language Models for Multi-Step Arithmetic Reasoning* by Jake O'Grady, Asena Isik Gürhan, Chee Fong Ting and Effirul Ramlan (University of Galway). We generated 20,000 GSM8K-derived arithmetic problems with step-by-step s...

O'Grady, Jake, Gürhan, Asena Isik, Chee, Fong Ting et al. · 465 citations
#computer vision Open access Jun 2016

Software Development in Startup Companies: The Greenfield Startup Model

The results are packaged in the Greenfield Startup Model (GSM), which explains the priority of startups to release the product as quickly as possible, and the need to shorten time-to-market, by speeding up the development through low-precision engineering activities.

Carmine Giardino, Nicolò Paternoster, M. Unterkalmsteiner et al. · 178 citations · ⚡14
#computer vision Open access Oct 2016

Software Startups - A Research Agenda

Software startup companies develop innovative, software-intensive products within limited timeframes and with few resources, searching for sustainable and scalable business models.

M. Unterkalmsteiner, P. Abrahamsson, Xiaofeng Wang et al. · 157 citations · ⚡17
#machine learning Review Open access Oct 2016

“Failures” to be celebrated: an analysis of major pivots of software startups

This study conducts a case survey study based on the secondary data of the major pivots happened in 49 software startups, and demonstrates that customer need pivot is the most common among all pivot types.

Sohaib Shahid Bajwa, Xiaofeng Wang, Anh Nguyen-Duc et al. · 127 citations · ⚡15
#computer vision Review Open access May 2015

A survey study on major technical barriers affecting the decision to adopt cloud services

The comparison of adopter and non-adopter sample reveals three potential adoption inhibitor, security, data privacy, and portability, which underlines the importance of the technical and security perspectives for research investigating the adoption of technology.

Nattakarn Phaphoom, Xiaofeng Wang, S. Samuel et al. · 111 citations · ⚡8
#computer vision Open access Feb 2018

Lean Internal Startups for Software Product Innovation in Large Companies: Enablers and Inhibitors

This study investigates how Lean internal startup facilitates software product innovation in large companies and identifies its enablers and inhibitors, and shows the potential of the method-in-action framework to investigate the Lean startup approach in non-startup context.

Henry Edison, Nina M. Smørsgård, Xiaofeng Wang et al. · 78 citations · ⚡6

Related blog posts

Microsoft Research Blog Oct 12, 2026

Simulating student learning for better AI tutoring

StudentSim is a research system for evaluating and improving AI tutoring tools before they are tested with learners. It models how individuals may engage with tutors, giving feedback to improve tutor performance across subjects like chess, writing, and mathematics. The post Simulating student learning for better AI tutoring appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.