Unverified50% confidenceBenchmarkExact time
Cortex proposes "memory consolidation": distilling the gains brought by context into low-rank or FFN parameter updates merged into the model itself, recovering 88.9% of the context gains after 50 sequential fact edits while retaining 99.8% of the original performance.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedEach time a memory is activated, it briefly enters an unstable state (the reconsolidation process), during which it is susceptible to being rewritten by new information.71% similarUnverifiedFine-tuning and persistent memory solve different problems: fine-tuning changes model behavior through training samples, while persistent memory lets the application store information externally and retrieve it when needed.70% similarUnverifiedThe performance gains of the forgetting mechanism were achieved under the same compute budget, indicating that it changes how effectively the investment is used rather than how much is invested.70% similarUnverified结合情境线索的词汇记忆效率可提升40%-60%70% similarUnverifiedIn FSRS, Stability describes the number of days required for memory retention to reach 90%, and Retrievability is the probability of successful recall at the current moment.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1067953API
curl https://kongchang.com/api/v1/knowledge/claims/1067953MCP
get_claim(id=1067953)