<p>Tail Spreading is a concept LoRA trained to create tail-assisted spreading poses and interactions while preserving character quality, anatomy, and overall composition.</p><p>Designed for anime-focused checkpoints, this LoRA teaches the relationship between a character's tail and pose, allowing the tail to become an active part of the scene rather than a simple background feature.</p><h2 id="features">Features</h2><ul><li><p>Tail-assisted spreading poses</p></li><li><p>Natural tail interaction and positioning</p></li><li><p>Works with a variety of tail types</p></li><li><p>Supports multiple camera angles and compositions</p></li><li><p>Compatible with demon girls, monster girls, kemonomimi, fox girls, wolf girls, dragon girls, and other tailed characters</p></li><li><p>Maintains character identity and scene composition</p></li><li><p>Trained on Illustrious XL</p></li></ul><h2 id="trigger-word">Trigger Word</h2><pre><code>tail_spreading</code></pre><h2 id="recommended-weight">Recommended Weight</h2><pre><code>0.8 - 1.2</code></pre><h2 id="usage-notes">Usage Notes</h2><ul><li><p>Best results are achieved when the character has a clearly visible tail.</p></li><li><p>Works particularly well with demon, monster, fantasy, and kemonomimi characters.</p></li><li><p>Can be combined with character, lighting, environment, anatomy, and style LoRAs.</p></li><li><p>Different poses and camera angles will naturally produce different tail interactions.</p></li><li><p>Higher weights generally increase concept strength and tail involvement.</p></li></ul><h2 id="example-prompt">Example Prompt</h2><pre><code>1girl, solo, demon girl, black hair, red eyes, tail, tail_spreading, kneeling, bent_over, looking_at_viewer,</code></pre><h2 id="example-prompt">Example Prompt</h2><pre><code>1girl, solo, fox girl, tail_spreading, on_back, spread_legs, looking_at_viewer, bedroom, </code></pre><h2 id="training-information">Training Information</h2><p>Base Model: Illustrious XL</p><p>Type: Concept LoRA</p><p>Purpose: Tail-assisted pose and interaction generation.</p><p>This LoRA was trained on a diverse dataset containing multiple characters, tail types, body types, camera angles, and compositions to maximize flexibility and encourage strong concept generalization across a wide variety of prompts.</p><p>If you create images using this LoRA, feel free to share them in the gallery. Feedback, examples, and prompt discoveries are always appreciated.</p>

Files with SHA256: 703b6f2db74e4fc79a278402d5e7db55f4765278456f46c2dc8c504278b16475