A Deafening Silence: Catastrophic Forgetting Lives in the Output Embeddings of Tokens the Data Never Speaks

arXiv:2610.09835v1 Announce Type: cross Abstract: Continual pre-training and fine-tuning in Large Language Models (LLMs) inevitably induce catastrophic forgetting, typically mitigated by replay using often-inaccessible original data.

← Back to all developments · Signals are a development reported by more than one source. See the methodology.