---
name: deepdream-psychedelia
description: "Apply the DeepDream Psychedelia style (2015–2017) when generating web pages, apps, memes, short video frames, forum and chat screens, and other images of internet culture. Load when the user asks for: deepdream, inceptionism, neural hallucination, dog faces in clouds, psychedelic AI, puppy slug."
domain: internet
era: feed-mobile
triggers:
  - "deepdream"
  - "inceptionism"
  - "neural hallucination"
  - "dog faces in clouds"
  - "psychedelic AI"
  - "puppy slug"
vector_profile:
  ornamentation: 5
  dimensionality: 3
  grid_rigidity: 1
  skeuomorphic_realism: 2
  typographic_expressiveness: 2
---

# DeepDream Psychedelia

> A machine that looks for patterns finds them in everything: the picture shows what the network expects to see, and it shows it everywhere.

Photographs that a neural network fills with eyes, dog faces, pagodas and swirled fur at every scale. It started in 2015 when Google released the code that made the effect.

**Period:** 2015–2017 · **Era:** The feed and the phone · **Domain:** Internet · **Also called:** DeepDream, Inceptionism, Puppy-slug dreams

## When to load this skill

Load it when the user asks for: "deepdream", "inceptionism", "neural hallucination", "dog faces in clouds", "psychedelic AI", "puppy slug".

Good for: music visuals and album art, posters on machine perception, explainers about how neural networks see, experimental art pages.

Do not use it when:
- The page needs a calm, trustworthy tone.
- The content is dense text that must be read at length.
- The audience is sensitive to repeated eye motifs.

## System directives

When generating web pages, apps, memes, short video frames, forum and chat screens, and other images of internet culture in DeepDream Psychedelia, you MUST:

1. Fill the frame with repeated motifs: eyes, snouts, feathers and arches at 3 to 5 sizes, from 8px to 200px.
2. Run 4 to 8 octaves of the effect, so that small motifs sit inside large ones.
3. Set saturation at 130% to 180% and push the hue toward orange, lime and magenta over a dark ground (#14102A).
4. Bend lines into spirals and wave fields; keep no straight edge longer than 10% of the frame.
5. Keep one readable photographic base, such as a landscape or a face, at about 30% visibility under the motifs.
6. Place a plain 16px mono caption bar at the foot of the frame with the layer name, for example inception_4c.
7. Loop the zoom at 12 fps: scale the image 1.03x for each frame and feed it back to the network.

## Guardrails

Each line below breaks the style. Check your output against the list before you return it.

- **NEVER:** Clean flat colour fields with no repeated motif.
- **NEVER:** Straight, geometric layouts with wide margins.
- **NEVER:** Photoreal skin and soft studio light.
- **NEVER:** Calm monochrome palettes.
- **NEVER:** One focal subject on an empty ground.
- **NEVER:** Neutral corporate type.

## Vector profile

| Axis | Score | 1 means | 5 means |
|---|---|---|---|
| Ornamentation | `■■■■■` 5/5 | pure minimal | maximalist |
| Dimensionality | `■■■□□` 3/5 | strictly 2D | full 3D / spatial |
| Grid rigidity | `■□□□□` 1/5 | chaotic / organic | strict mathematical grid |
| Skeuomorphic realism | `■■□□□` 2/5 | abstract digital | photorealistic material |
| Typographic expressiveness | `■■□□□` 2/5 | neutral / invisible | experimental / illustrative |

Hold these scores. If a request moves an axis by 2 points or more, you have left the style: go to "Lineage fallbacks".

## Spatial & grid logic

- **Grid mechanics:** A full-bleed image with a 12-column caption area of 1 row; the image runs to the edge and the frame has no gutter.
- **Whitespace ratio:** 0% to 5% of the image stays empty; the caption bar takes 48px and the spacing scale is 8, 16 and 32px.
- **Alignment rules:** Flush-left caption under a centred, full-width image; no element is centred on a motif.
- **Container geometry:** Square or 4:3 frames with a 0px radius; caption chips with a 4px radius.

## Typography

- **Classifications:** monospace · humanist sans
- **Tracking:** Display takes +0.02em; captions take +0.04em.
- **Leading:** Display 1.15; body 1.5.
- **Hierarchy rules:** Use 2 weights: regular (400) and bold (700). · Use sizes 12, 16, 28 and 56px. · Set layer names in lower case with underscores. · Keep the headline short: 3 words at most.

**Font pairings**

| Display | Body | Note |
|---|---|---|
| Space Mono | IBM Plex Mono | Mono faces that suit layer names and code captions. |
| Syne | Inter | A wide display face with a neutral text face for page copy. |

## Colour & lighting

- **Palette structure:** Four saturated hues (orange, lime, magenta, cyan) on a deep indigo ground, with one cream highlight.
- **Contrast rules:** Cream text on indigo holds at least 12:1; the saturated hues carry image fields and borders, not long text.
- **Surface treatment:** High-saturation texture fields of fine, repeated detail; a 3% grain; no flat fills.
- **Light and shadow:** No cast shadow; each motif has its own inner glow of 2px to 6px and a rim of lighter hue.

**Palette**

| Name | Hex | Role |
|---|---|---|
| Dream Indigo | `#14102A` | background |
| Eye Orange | `#FF9A3C` | primary |
| Acid Lime | `#9BE564` | accent |
| Dream Magenta | `#E83E8C` | second accent |
| Pool Cyan | `#3EC7E8` | highlight |
| Fur Cream | `#F6E7C1` | text |

## Materiality & texture

- **Surface simulation:** Over-fitted feature maps: fur, scales, feathers and eyes that grow from the texture of the photograph.
- **Border rules:** A 1px border in lime at 60% on caption chips; the image has no border.
- **Blur and elevation:** No blur and no elevation; detail is sharp at every scale.

## Ornamentation & iconography

- **Iconography:** Outline icons at 20px with a 2px round stroke; used only in the caption bar.
- **Illustration:** Repeated eyes, dog snouts, birds, pagodas and arches that grow out of clouds, trees and faces.
- **Decorative elements:** repeated eye motif · spiral fur · layer-name caption · zoom-loop frame · octave tiling

## Motion & interactivity

- **Born on screen:** yes
- **Transition physics:** An endless zoom: each frame scales 1.03x and adds new motifs; the loop runs 8 to 12 seconds at 12 fps.
- **Easing:** cubic-bezier(0.37, 0, 0.63, 1)
- **Duration (ms):** 600
- **Hover and active:** Hover raises saturation by 20% and adds one octave of motifs; active freezes the zoom.

## Code & token primer

Set `data-style="deepdream-psychedelia"` on the root element and paste the tokens. The file [`tokens.css`](./tokens.css) has the same tokens and a set of reference rules.

```css
[data-style="deepdream-psychedelia"] {
  --color-bg: #14102A;
  --color-surface: #231A47;
  --color-text: #F6E7C1;
  --color-muted: #CBBFE8;
  --color-primary: #FF9A3C;
  --color-accent: #9BE564;
  --color-border: #9BE564;
  --font-sans: "IBM Plex Mono", "Inter", monospace;
  --font-display: "Space Mono", "Syne", monospace;
  --border-radius: 4px;
  --border-width: 1px;
  --shadow-elevation: 0 0 12px rgba(255,154,60,0.45);
  --backdrop-filter: none;
  --space-unit: 8px;
  --ease: cubic-bezier(0.37, 0, 0.63, 1);
  --duration: 600ms;
  --type-ratio: 1.5;
  --size-small: 10.67px;
  --size-h3: 36px;
  --size-h2: 54px;
  --size-h1: 81px;
  --color-on-bg: #F6E7C1;
  --color-on-surface: #F6E7C1;
  --color-on-primary: #14102A;
  --color-on-accent: #14102A;
  --color-muted-on-bg: #CBBFE8;
  --color-muted-on-surface: #CBBFE8;
  --color-link: #9BE564;
  --color-link-on-surface: #9BE564;
  --display-weight: 700;
  --subhead-weight: 400;
  --body-weight: 400;
  --strong-weight: 700;
  --label-weight: 500;
  --display-tracking: 0.02em;
  --display-leading: 1.15;
  --display-case: none;
  --body-leading: 1.5;
  --body-size: 16px;
  --icon-stroke: 2;
  --ground: linear-gradient(100deg, var(--color-bg) 26%, transparent 58%),
    repeating-conic-gradient(from 0deg at 82% 62%, var(--color-primary) 0 12deg, var(--color-accent) 12deg 24deg, var(--color-primary) 24deg 36deg),
    var(--color-bg);
}
```

Tailwind patterns:

| Element | Classes |
|---|---|
| page | `bg-[#14102A] text-[#F6E7C1] font-mono antialiased px-4 md:px-10` |
| heading | `text-3xl md:text-6xl font-bold leading-[1.15] text-[#FF9A3C]` |
| body | `text-base leading-[1.5] max-w-prose` |
| section | `rounded border border-[#9BE564]/60 bg-[#231A47] shadow-[0_0_12px_rgba(255,154,60,0.45)]` |
| button | `rounded border border-[#9BE564] bg-[#FF9A3C] px-5 py-3 text-sm font-bold text-[#14102A] transition-all duration-[600ms] hover:saturate-150` |
| label | `text-xs lowercase tracking-[0.04em] text-[#9BE564]` |

## Visual pack

The pictures and icons of this style are files. Use them as they are, or read them as a specification and draw your own in the same language.

| File | Content |
|---|---|
| [`pack/hero.svg`](./pack/hero.svg) | Hero plate, 800 by 600. An abstract picture in the visual language of the style. |
| [`pack/pattern.svg`](./pack/pattern.svg) | Pattern tile, 80 by 80. It repeats in both directions. |
| [`pack/icons.svg`](./pack/icons.svg) | Icon sprite: home, search, heart, star, bell, user, mail, calendar, image, bolt, play, check, close, sliders. |

- **Icons:** outline, stroke 2 on a 24px field, round caps, round joins, container: none. Use: `<svg class="icon"><use href="pack/icons.svg#home"/></svg>`. An icon takes `currentColor`.
- **Icon family:** the sprite is cut from Phosphor (Duotone) by Phosphor Icons, licence MIT. Reason: A duotone family that keeps the caption icons light on a very busy image. It is an open-source stand-in for a demo. For a product, use the icon library of the brand and choose the family by the rules below.
- **Display type:** weight 700, tracking 0.02em, leading 1.15, case: none.
- **Weights:** display 700, subhead 400, body 400, bold in text 700, labels and buttons 500. The style uses 3 weights (400, 500, 700) and no other. The display is 300 heavier than the body. The accent is colour: one coloured word or field is the accent, so the weights stay quiet.
- **Body type:** 16px, leading 1.5.
- **Page:** hero poster, center aligned, compact density, accent tilt 0°.
- **Ground:** swirl. The ground repeats the spiral fur and the wave fields that DeepDream grows from a photograph. Put the class `ground` on the first screen of the page; it takes `--ground`, a stack of gradients built from the colour tokens. The ground can run behind the copy, so put the class `ground-copy` on each text block of that screen: the block gets `--color-bg` and a soft edge. Keep the rest of the page on `--color-bg`.
- **Text colour:** `--color-text` is for the main reading field only. On each field, use its token: `--color-on-bg`, `--color-on-surface`, `--color-on-primary`, `--color-on-accent`.

### How to choose an icon pack

An icon pack is a typeface for pictures. Choose it with the same care as the text face. These rules hold for an open-source set and for a licensed library.

1. **The stroke matches the stem.** Set an icon next to the body text at text size. The line of the icon and the stem of a letter have the same weight.
2. **Terminals and corners follow the radius token.** Butt caps and mitre joins go with a radius of zero. Round caps and round joins go with a soft radius.
3. **The fill mode follows the technology of the era.** Print eras cut solid forms. Screen eras with depth fill and shade the form. Current product interfaces draw an outline.
4. **The metaphors are not younger than the style.** Each picture is an object that the era knew. An icon with a later metaphor breaks the period.
5. **One family, one grid.** All icons come from one family and one variant, drawn on one field. Do not mix families on one surface.
6. **The family covers the product.** List each concept that the product needs before you commit. A family that has no icon for one of them is the wrong family.
7. **The colour is a token.** An icon takes currentColor. A second tone comes from a token, not from the file.
8. **The licence permits the use.** Read the licence of the set. Give credit where the licence asks for it. A product with a brand uses its licensed library and applies the same rules to choose the family.

## Techniques

These techniques started in one domain and moved into this style. Apply the rules as they are written; the full record is in [`TECHNIQUES.md`](../TECHNIQUES.md).

### Feature amplification (process)

A method that changes a photograph to raise the response of one network layer, so the network patterns grow out of the image. Origin: internet, Google, 2015.

- Pick one layer of a trained image network; a middle layer gives eyes and fur, a high layer gives whole animals.
- Change the pixels of the image by small steps so that the layer response rises.
- Repeat at 4 to 8 image sizes (octaves) and mix the results back into one image.
- Feed the output back in and zoom 3% for each frame to make an endless loop.

## Organisations

The groups, studios, companies, schools, channels and backers that carried this style.

- **GitHub** (platform, web, 2007–now), published the style. A web site where developers host code. Its own dark interface and the editor themes and terminal tools that it hosts set the look of software for engineers.

## Permutations

Each knob varies the style inside a range where it stays itself. Keep each value in its range. `npm run permute -- <id> key=value` writes the tokens for other values.

| Knob | Default | Range | Changes |
|---|---|---|---|
| Base colour (`base_colour`) | #FF9A3C | #FF9A3C, #14102A, #231A47, #0B0818 | --color-primary, --color-on-primary, --color-link-on-surface |
| Accent colour (`accent_colour`) | #9BE564 | #FF9A3C, #9BE564, #E83E8C, #3EC7E8 | --color-accent, --color-on-accent, --color-link |
| Corner radius (`corner_radius`) | 4px | 0–8 px | --border-radius |
| Grid step (`grid_step`) | 8px | 4–16 px | --space-unit, --grid-gutter |
| Type scale ratio (`type_scale_ratio`) | 1.5 | 1.25–1.618 | --type-ratio, --size-small, --size-h3, --size-h2, --size-h1 |

## Lineage fallbacks

When the user asks for a variation, do not improvise. Blend toward the neighbour named here, and load its skill.

- If the user asks for **calmer**: move toward [`surrealism-web`](../surrealism-web/SKILL.md) (Surrealism Web). Keep one dream object on an empty ground and cut saturation to 80%; remove the repeated motifs.
- If the user asks for **more recent**: move toward [`diffusion-gloss`](../diffusion-gloss/SKILL.md) (Diffusion Gloss). Replace the motif swirl with one smooth, over-detailed render and add a teal and amber rim light.
- If the user asks for **more retro**: move toward [`psychedelic`](../psychedelic/SKILL.md) (Psychedelic). Swap the machine motifs for hand-drawn paisley and flower forms with a halftone ground.

- **Led to (go here for a newer version):** [`deepfake-face-swap`](../deepfake-face-swap/SKILL.md), [`gan-portrait`](../gan-portrait/SKILL.md)
- **Kin in another domain:** [`psychedelic`](../psychedelic/SKILL.md), [`psychedelic-concert-poster`](../psychedelic-concert-poster/SKILL.md)
- **Nearest by vector profile:** [`seapunk`](../seapunk/SKILL.md), [`tumblr-soft-grunge`](../tumblr-soft-grunge/SKILL.md), [`ai-avatar-portrait`](../ai-avatar-portrait/SKILL.md)

## Emotional target

The viewer feels awe and unease: a familiar picture is full of watchers.

**Keywords:** hallucinatory, dense, uncanny, organic, saturated, recursive · **Brand archetypes:** magician, explorer, jester

## Sources

- [Wikipedia: DeepDream](https://en.wikipedia.org/wiki/DeepDream)
- [Google Research Blog: Inceptionism, Going Deeper into Neural Networks (2015)](https://research.google/blog/inceptionism-going-deeper-into-neural-networks/)

---

Generated from [`data/internet/styles/deepdream-psychedelia.json`](../../data/internet/styles/deepdream-psychedelia.json) by [design-ontology](../../README.md). Do not edit this file; change the data and run `npm run build`.
