Verified70% confidenceFactExact time
The core challenge of internalizing knowledge into model parameters is the Catastrophic Forgetting problem, which is a difficulty not yet fully resolved in the field of continual learning.
4
Sources
70%
Confidence
Long-term
Relevance
5/16/2026
First Seen
Sources
memU Memory Framework Explained: Unifying Multi-Modal Agent Memory with a File System
bilibili日新月异max5/16/2026
Related Entities
Related Claims
UnverifiedThe core challenge of continual learning is catastrophic forgetting, which Liang Wenfeng considers a bottleneck that has not yet been solved anywhere in the world.82% similarUnverified专项优化的AI模型在特定领域可能表现不及通用基础模型，这与深度学习中的灾难性遗忘现象有关79% similarVerifiedThe core capabilities of large language models are reasoning and question answering, but they lack memory and their knowledge has a temporal cutoff at the time of training.76% similarUnverified机器遗忘（machine unlearning）在技术上仍是未解决难题——从已训练好的模型中精确删除特定数据的影响目前尚无可靠方法74% similarUnverifiedThe study reveals a knowing-doing gap: improvements in a model's knowledge memorization cannot reliably generalize to improvements in clinical reasoning ability.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1566API
curl https://kongchang.com/api/v1/knowledge/claims/1566MCP
get_claim(id=1566)