Skip to content

Commit 27cf6ef

Browse files
committed
update
1 parent 3506023 commit 27cf6ef

32 files changed

Lines changed: 764 additions & 388 deletions

.codex/hooks.json

Lines changed: 17 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,17 @@
1+
{
2+
"hooks": {
3+
"PostToolUse": [
4+
{
5+
"matcher": "Edit|Write|apply_patch",
6+
"hooks": [
7+
{
8+
"type": "command",
9+
"command": "node \"/Users/yjw0510/.agents/skills/impeccable/scripts/hook.mjs\"",
10+
"timeout": 5,
11+
"statusMessage": "Checking UI changes"
12+
}
13+
]
14+
}
15+
]
16+
}
17+
}

DESIGN.md

Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -229,6 +229,25 @@ fallback gradient stops `rgba(17,24,39,.92)` `rgba(5,5,16,.98)`
229229
`rgba(241,245,249,.9)`. CPK/element colors inside WebGL scenes are data,
230230
not design tokens.
231231

232+
Schematic accent tokens (`--sch-*` in globals.css): every chromatic accent a
233+
schematic animates (SCF ring, packets, force-field term colors, electron
234+
density) is declared as a light/dark pair. Hue is held across modes; dark
235+
raises lightness one step so animated strokes read on the instrument ground.
236+
Light: `--sch-amber #d97706` `--sch-amber-bright #f59e0b` `--sch-amber-label
237+
#92400e` `--sch-density #1d4ed8` `--sch-stretch #0891b2` `--sch-angle #16a34a`
238+
`--sch-lj #7e22ce` `--sch-coulomb #4f46e5` `--sch-delta-plus #2563eb`
239+
`--sch-delta-minus #c026d3` `--sch-packet #c41e3a` `--sch-peripheral #475569`
240+
`--sch-meso-n #7c8db0` `--sch-meso-o #b07c7c`. Dark: `#f59e0b` `#fbbf24`
241+
`#fbbf24` `#60a5fa` `#22d3ee` `#4ade80` `#c084fc` `#818cf8` `#60a5fa`
242+
`#e879f9` `#fb7185` `#94a3b8` `#9fb0d0` `#cf9d9d`. Mode-stable scientific
243+
literals (CPK-adjacent mid tones, HOMO/LUMO, frame hairlines) stay inline:
244+
`#ef4444` `#dc2626` `#f1f5f9` `#3b82f6` `#06b6d4` `#f97316` `#9ca3af`
245+
`#cbd5e1`. MLFF force-arrow magnitude interpolants (data, `#3b82f6``#ef4444`
246+
per-atom): `#3e80f2` `#3f80f1` `#427fee` `#497ce7` `#4b7ce5` `#5678da`
247+
`#5e75d2` `#6374ce` `#6573cc` `#6d70c4` `#6f6fc1` `#786cb9` `#97629a`
248+
`#9f5f92` `#a05f91` `#a55d8c` `#a75c8a` `#aa5b88` `#ac5a85` `#b7577a`
249+
`#bb5576` `#c15370`.
250+
232251
## Anti-slop enforcement (binding)
233252

234253
Zero em/en-dashes in visible copy (Korean gains 없음: use `,`/`·`/period —

content/multiscale/allatom.mdx

Lines changed: 13 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,11 @@
11
---
22
title: "All-Atom Molecular Dynamics"
33
titleKo: "전원자 분자동역학"
4-
shortDescription: "Retaining full atomic resolution without electronic structure calculations, this method describes interatomic interactions to capture both static distributions and dynamic properties of atoms and molecules"
5-
shortDescriptionKo: "개별 원자의 해상도를 유지하되 전자구조는 계산하지 않는다. 원자 수준에서 상호작용을 기술해 원자와 분자의 정적인 분포와 동적인 특성을 포착하는 기법이다"
4+
shortDescription: "Tracks explicit atoms and solvent with a classical potential, providing trajectory-level structural and dynamical references for larger-scale models."
5+
shortDescriptionKo: "고전 퍼텐셜로 명시적 원자와 용매를 추적해, 더 큰 스케일 모형의 구조 및 동역학 참조값을 제공한다."
66
icon: "Atom"
77
color: "#06b6d4"
8-
order: 2
8+
order: 3
99
scale: "1-10 nm"
1010
contentKo: |
1111
## 개요
@@ -21,7 +21,11 @@ contentKo: |
2121
2222
## 방법론
2323
24-
이 페이지의 시각화는 물에 녹인 카페인의 명시적 용매 궤적을 사용하여 전원자 MD의 핵심 개념을 예시한다: 원자 표현 방식, 힘장 항이 실제로 어떻게 보이는지, 국소 분자 구조가 장소에 따라 어떻게 달라지는지, 주기 경계 조건이 어떻게 작동하는지. 이 궤적은 하나의 예시이며, 이 개념들은 액체나 용액의 모든 전원자 시뮬레이션에 적용된다.
24+
멀티스케일 인터랙티브 경험은 물에 녹인 카페인의 명시적 용매 궤적으로 전원자 MD의 핵심 개념을 예시한다. 원자 표현 방식, 힘장 항, 국소 분자 구조의 차이, 주기 경계 조건을 차례로 보여 준다. 이 궤적은 하나의 학습용 예시이며, 핵심 개념은 액체와 용액의 다양한 전원자 시뮬레이션에 적용된다.
25+
26+
## 멀티스케일 체계에서의 역할
27+
28+
전원자 MD는 더 정교한 전자구조 또는 MLFF 계산과 더 단순한 조대화 모형 사이에서 구조와 동역학의 기준선을 제공한다. 같은 계의 분포, 확산, 접촉 통계를 비교하면 각 해상도에서 무엇을 보존했고 무엇을 생략했는지 확인할 수 있다.
2529
2630
## 관련 논문
2731
@@ -44,7 +48,11 @@ All-atom molecular dynamics tracks every atom in a system, including every water
4448

4549
## Methods
4650

47-
The interactive visualizations on this page use an explicit-solvent trajectory of caffeine dissolved in water to illustrate core all-atom MD concepts: how atoms are represented, what force-field terms look like in practice, how local molecular structure varies from place to place, and how periodic boundary conditions work. The trajectory is one example; these concepts apply to any all-atom simulation of a liquid or solution.
51+
The Multiscale interactive experience uses an explicit-solvent trajectory of caffeine dissolved in water to illustrate core all-atom MD concepts: atomistic representation, force-field terms, local molecular structure, and periodic boundary conditions. The trajectory is a teaching example; the concepts apply across all-atom simulations of liquids and solutions.
52+
53+
## Role in Multiscale Framework
54+
55+
All-atom MD provides structural and dynamical reference measurements between more detailed electronic-structure or MLFF calculations and simpler coarse-grained models. Comparing distributions, diffusion, and contact statistics for the same system makes clear which features are retained or omitted at each resolution.
4856

4957
## Related Publications
5058

content/multiscale/dft.mdx

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,11 @@
11
---
22
title: "First-Principles Calculations (DFT)"
33
titleKo: "제일원리 계산 (DFT)"
4-
shortDescription: "Rigorously computing electronic structure at the quantum-mechanical level to obtain properties such as band gaps, and to determine interatomic interactions and molecular geometries with the highest fidelity"
5-
shortDescriptionKo: "양자역학 수준에서 전자구조를 엄밀하게 계산하여 밴드 갭 같은 성질을 얻어내고, 원자 간 상호작용과 분자 구조를 가장 높은 정밀도로 규명한다"
4+
shortDescription: "Uses quantum-mechanical electronic-structure calculations to generate reference energies, forces, and geometries. The selected approximation and dispersion treatment determine how results are assessed."
5+
shortDescriptionKo: "양자역학적 전자구조 계산으로 에너지, 힘, 구조의 참조값을 만들며, 선택한 근사와 분산 보정에 따라 결과를 검토한다."
66
icon: "Orbit"
77
color: "#f97316"
8-
order: 4
8+
order: 1
99
scale: "0.1-1 nm"
1010
contentKo: |
1111
## 개요

content/multiscale/meso.mdx

Lines changed: 11 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,11 @@
11
---
22
title: "Mesoscale Simulation"
33
titleKo: "메조스케일 시뮬레이션"
4-
shortDescription: "Coarse-graining simplifies chemical detail to reduce computational cost, enabling the study of nanomaterials at larger length and longer time scales than other methods"
5-
shortDescriptionKo: "조대화로 화학적 세부를 단순화해 계산 비용을 낮추고, 다른 방법보다 더 크고 긴 스케일에서 나노 재료를 연구한다"
4+
shortDescription: "Uses coarse-grained representations to connect molecular interactions to collective behavior at larger length and time scales."
5+
shortDescriptionKo: "조대화 표현으로 분자 상호작용을 더 큰 길이 및 시간 스케일의 집단 거동과 연결한다."
66
icon: "Layers"
77
color: "#f59e0b"
8-
order: 1
8+
order: 4
99
scale: "10 nm - 1 μm"
1010
contentKo: |
1111
## 개요
@@ -23,6 +23,10 @@ contentKo: |
2323
2424
전원자 시뮬레이션과 실험 산란·현미경 데이터(SAXS, TEM, AFM)에 견주어 조대화 모형을 개발하고 검증한다. 이 모형으로 수백만 상호작용점을 담은 계를 마이크로초 규모로 시뮬레이션한다.
2525
26+
## 멀티스케일 체계에서의 역할
27+
28+
메조스케일은 앞선 해상도에서 정리한 상호작용과 구조적 제약을 집단적 거동으로 확장한다. 여기서는 정확한 원자 배치 하나보다 분포, 자기조립 경로, 시간에 따른 형태 변화가 더 중요한 검증 대상이 되며, 산란과 현미경 관측이 그 비교 기준을 제공한다.
29+
2630
## 관련 논문
2731
2832
- Yu et al., *Advanced Science* (2024), 접합 고분자의 이중 영역 배좌
@@ -46,6 +50,10 @@ Mesoscale simulations bridge the gap between atomistic methods (accurate but com
4650

4751
We develop and validate coarse-grained models against all-atom simulations and experimental scattering and microscopy data (SAXS, TEM, AFM). These models simulate systems of millions of interaction sites over microsecond timescales.
4852

53+
## Role in Multiscale Framework
54+
55+
Mesoscale simulation carries the interactions and structural constraints established at earlier resolutions into collective behavior. Here, distributions, self-assembly pathways, and time-dependent morphology matter more than one exact atomic arrangement. Scattering and microscopy provide the comparison frame.
56+
4957
## Related Publications
5058

5159
- Yu et al., *Advanced Science* (2024), Two-regime conformation of grafted polymers

content/multiscale/mlff.mdx

Lines changed: 13 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -1,11 +1,11 @@
11
---
22
title: "Machine Learning Force Fields"
33
titleKo: "머신러닝 역장"
4-
shortDescription: "Machine learning force fields achieve quantum-level accuracy at near-classical MD speed, tackling problems inaccessible to conventional force fields"
5-
shortDescriptionKo: "머신러닝 역장은 양자 수준 정확도를 고전 MD에 버금가는 속도로 구현하는 계산화학 기법으로, 기존 힘장으로는 접근하기 힘들었던 난제들을 공략한다"
4+
shortDescription: "Learns an interatomic potential from quantum-mechanical reference data, aiming to retain near-reference accuracy at lower cost for the atomic environments represented in training."
5+
shortDescriptionKo: "양자역학 참조 데이터로 원자 간 퍼텐셜을 학습해, 학습에 포함된 원자 환경에서 더 낮은 비용으로 참조 수준에 가까운 정확도를 목표로 한다."
66
icon: "Brain"
77
color: "#8b5cf6"
8-
order: 3
8+
order: 2
99
scale: "1-10 nm (ab initio accuracy)"
1010
contentKo: |
1111
## 개요
@@ -21,7 +21,11 @@ contentKo: |
2121
2222
## 방법론
2323
24-
밀도범함수이론(DFT) 양자역학 참조 데이터로 MLFF를 훈련하며, 물리적 대칭을 만족하는 신경망 아키텍처를 사용한다. 분자를 회전하거나 이동해도 예측이 변하지 않는다. 희귀 사건의 관찰을 가속하는 향상된 표본추출 기법과 결합하면 양자에 가까운 정확도로 나노초 규모의 시뮬레이션을 수행할 수 있다.
24+
밀도범함수이론(DFT) 양자역학 참조 데이터로 MLFF를 훈련하며, 물리적 대칭을 만족하는 신경망 아키텍처를 사용한다. 분자를 회전하거나 이동해도 예측이 변하지 않는다. 향상된 표본추출 기법과 결합하면, 학습에 포함된 원자 환경에서 참조 수준에 가까운 정확도를 목표로 더 긴 시간 규모를 탐색할 수 있다. 적용 범위는 별도의 참조 데이터로 계속 점검한다.
25+
26+
## 멀티스케일 체계에서의 역할
27+
28+
MLFF는 DFT가 만든 에너지와 힘의 참조값을 더 긴 원자 수준 궤적에 쓸 수 있는 퍼텐셜로 바꾼다. 이 연결은 전자구조 계산만으로는 다루기 어려운 시간과 계 크기를 탐색하게 해 주며, 전원자 및 조대화 모형과 비교할 때는 동일한 구조·동역학 측정량으로 적용 범위를 점검한다.
2529
2630
## 관련 논문
2731
@@ -43,7 +47,11 @@ In molecular simulation, a force field is the mathematical model that describes
4347

4448
## Methods
4549

46-
We train MLFFs on quantum-mechanical reference data from density functional theory (DFT) calculations using neural network architectures that respect physical symmetries; predictions remain unchanged when the molecule is rotated or translated. Combined with enhanced sampling techniques that accelerate the observation of rare events, these force fields reach nanosecond simulation time scales with near-quantum accuracy.
50+
We train MLFFs on quantum-mechanical reference data from density functional theory (DFT) calculations using neural network architectures that respect physical symmetries; predictions remain unchanged when the molecule is rotated or translated. Combined with enhanced sampling techniques that accelerate the observation of rare events, they can explore longer time scales while targeting near-reference accuracy for atomic environments represented in training. Their domain of reliable use still requires checks against separate reference data.
51+
52+
## Role in Multiscale Framework
53+
54+
An MLFF turns DFT reference energies and forces into a potential that can be evaluated along longer atomistic trajectories. This connection opens time and system-size regimes that direct electronic-structure calculations cannot routinely reach. Its scope is checked with shared structural and dynamical measurements when it is compared with all-atom or coarse-grained models.
4755

4856
## Related Publications
4957

content/publications/2025-self-assembly-review.mdx

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -18,7 +18,7 @@ correspondingAuthors:
1818
- "Myungeun Seo"
1919
- "YongJoo Kim"
2020
tags:
21-
- "Review"
21+
- "Self-Assembly"
2222
featured: true
2323
---
2424

0 commit comments

Comments
 (0)