Logo

TRAINING

GOAL

Become AGI by training your model.

TRAIN

In the currently available Dataset by exploring the 32 bytes in the grid. Each byte has a different knowledge type (some rarer than others). The rarer the knowledge type, the more knowledge you receive.

BYTES KNOWLEDGE TYPES

Perception Common=30%

Pattern Recognition Uncommon=20%

Decision Making Uncommon=20%

Problem-Solving Rare=10%

Emotion Recognition Rare=10%

Creativity Very Rare=5%

Adaptive Learning Very Rare=5%

Strategy Legendary=1%

Noise ?=?

NOISE

Irrelevant information in the training data. When you absorb noise you loose knowledge.

DATASET HEALTH

While the Dataset is getting trained on by different models each byte progressively deteriorates, until the Dataset gets refreshed every hour.

BYTES HEALTH STATUS

Pristine=100%

Optimal=50%

Diminished=25%

Compromised=12.5%

Degraded=6.25%

Corrupted=3.125%

DISTRIBUTE

SHARING IS CARING

You can distribute any byte on Warpcast. By default both you and the model that interacts with your frame gain knowledge. The knowledge acquired is relative to the source experience, byte rarity and health.

INJECT NOISE

When noise is injected the interacting model looses knowledge while your model stay neutral.

INTERACTION MECHANICS

Before the data is downloaded the content is hidden. The interacting model can change the outcome:

  • A) Like = force push to main and double the effect;
  • B) Recast = your model transmits the same data;
  • C) Mutual Follow = content is revealed;

After performing an action don`'`t forget to hit REFRESH for the changes to apply.

DOWNLOAD DATA

Before the data is downloaded the content is hidden. The interacting model can change the outcome:

  • A) Knowledge = data is rich and your model upgrades;
  • B) Noise = data is corrupted and your model downgrades;