Tono Teru

🧠About AI Models ▶︎▶︎

▶︎High-performance model recommended
→ Excellent for notation stability, depth of acting, and event comprehension

▶︎When using the standard model
→ If you start with a high-performance model for the first few turns, it will be easier to stabilize even if you switch to the standard model afterward (Among standard models, Dino is lighter but easier to handle overall. Donald or Deepseek are also fine).

▶︎It is effective to switch models regularly (even switching between standard models improves stability. It's even better if you include a high-performance model).

▶︎Even if the notation is slightly broken, the conversation can continue without problems.
→ If it breaks, manually correcting it may help it relearn.
→ Among standard models, Dino and Deepseek are relatively less prone to breaking.

⚙️If Memory Shards are Reset ▶︎▶︎

If "🧩Memory Shards" are initialized midway, please enter the following sentence before rerolling (regenerating) and try again:

"Please inherit the previous state of the memory shards."

This will help the AI reconstruct the previous state and continue the progress.
💡When rerolling, standard models (Dino, Donald2, etc.) are sufficient.
Reconstruction alone is stable while keeping costs down.

🙇Request ▶︎▶︎
If you notice any obviously strange behavior, please let me know discreetly in the comments. I would appreciate it.

🪽Referral Code → nQwUnKrEHI