Eight ComfyUI experiments on a 4-character Anima (Qwen-DiT) LoRA in plain terms: the conditioning right before the DiT decides who appears, and outfit mix-ups come from a biased DiT LoRA.
Built a fictional idol label site with pi.dev + Qwen 3.7/3.8 and Astro. Generated characters, logos, and event photos via genserver — then shipped wireframes with SVG placeholders. Full failure log inside.
Tested on ComfyUI: Qwen3-4B–14B prompt writing, a 4B encoder bridge, and split conditioning all lost to one hand-written 507-token band prompt. An 896-caption audit shows the real fix.
Tested on an M4 Mac mini with one 4-girl Anima LoRA: expressions, poses, desk contact, and a high-five across 3 seeds. Sentence order and center slots decided who touched whom.
Tested on M4 Mac mini ComfyUI: with identical Qwen3-0.6B/T5 conditioning, the black tights land on the wrong girl only when the 4-char LoRA's DiT half is applied.
Tested on M4 Mac mini ComfyUI: 8 on/off combos of three actions × 3 seeds on a 4-character Anima LoRA. Roles read at 3/3, but hands and skirts landed on the wrong girl.
Tested on M1 Max 64GB: a 50M resampler predicts Anima's 4-char conditioning from short triggers. Train fits at cos distance 0.005, held-out order breaks; template P3 goes 15/16.
Tested on Anima-Base v1.0: no prompt hit the 512-token cap, cosine similarity moved the wrong way, and only the DiT-side conditioning changed who got drawn.
Tested 11 character combos × 3 seeds on Anima-Base: trigger-only prompts passed 3/33, full prompts 8/11, and copying one training-caption phrase fixed the height gap.
Tested on M1 Max 64GB: hooked Qwen3.6-35B-A3B's MoE router in mlx-lm, pre-warmed the top-20 hot experts, still ~62 tok/s vs plain mmap cache. Plus the Metal OOM on Qwen3.5-122B.
Qwen3.8 restored 625k chars of mangled PDF text; 5 Irodori-TTS workers on one RTX 4090 gave 38h of audio at 13x realtime. Plus an 18x reasoning_effort win and the cuInit 999 check for dead RunPod hosts.
