ERAIS technical claim
research.modular_expert_continual_learning.v1
Can new skills be learned without losing old ones?
Evidence maturity: BlockedSource-pinned record
Reviewed · neuroforge-io/erais
The wording and state below belong to this source snapshot. Related current context is separate from the recorded claim.
Approved wording · 12 July 2026
What the record says.
ERAIS is testing modular expert addition as a continual-learning strategy.
Required evidence
- Sequential donor addition, stratified replay, old-route and old-logit retention, rare-route coverage, OOD abstention, multiple seeds, and forgetting statistics.
Do not infer
- Do not claim continual learning without forgetting or effortless expert installation.
Continue in context
Related work.
Explore the current context. It does not update this recorded claim.
Read the research programmeRead the state together with its scope and evidence requirements. Commercial availability and technical maturity are separate decisions.
Return to the complete register