Prediction Dependency
A core metric used for contributor selection in RASO that characterizes a client's influence on the shared prediction block, commonly computed via cosine similarity or norm ratios.
Timeline (last 90 days)
RASO's contributor selection selects contributors based on prediction dependency and participation in local optimization.
Prediction dependency characterizes a client's influence on the shared prediction block and can be estimated by computing the cosine similarity or norm ratio between the client-uploaded prediction block gradients and the server-aggregated gradients, or by using prediction consistency on a validation set.
All Facts (2)
RASO's contributor selection selects contributors based on prediction dependency and participation in local optimization.
50%UnverifiedPrediction dependency characterizes a client's influence on the shared prediction block and can be estimated by computing the cosine similarity or norm ratio between the client-uploaded prediction block gradients and the server-aggregated gradients, or by using prediction consistency on a validation set.
50%