Indicators / series
evidence.continual_learning_unsolved.for.pt
Source: Google Research blog (Google) · Google; short quotation · CSV
Google Research
| id | text | grade | as of date | published date | tier | audited vs reported | extraction method | review status | retrieved at | http status | content hash | supersedes id | url | flags |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 4b623aa5074b1691 | Google Research introduces Nested Learning as a way round catastrophic forgetting: updating a model's weights with new data still costs it what it knew. | grade B, evidence tier 6 | 2025-11-07 | 2025-11-07 | 6 | reported | manual | approved | 2026-09-11 13:21 | 200 | 43c829b3c086 | — | source page |
Raw snippets and dispute text
- 4b623aa5074b1691 The simple approach, continually updating a model's parameters with new data, often leads to “catastrophic forgetting” (CF), where learning new tasks sacrifices proficiency on old tasks.