Mechanistic Analysis of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908788960788480 |
|---|---|
| author | Imanov, Olaf Yunus Laitinen |
| author_facet | Imanov, Olaf Yunus Laitinen |
| contents | Large language models exhibit remarkable performance across diverse tasks through pre-training and fine-tuning paradigms. However, continual fine-tuning on sequential tasks induces catastrophic forgetting, where newly acquired knowledge interferes with previously learned capabilities. Despite widespread observations of this phenomenon, the mechanistic understanding remains limited. Here, we present a comprehensive mechanistic analysis of catastrophic forgetting in transformer-based LLMs during sequential fine-tuning. Through systematic experiments across multiple model scales (109B to 400B total parameters) and task sequences, we identify three primary mechanisms driving forgetting: gradient interference in attention weights, representational drift in intermediate layers, and loss landscape flattening. We demonstrate that forgetting severity correlates strongly with task similarity (Pearson r = 0.87) and gradient alignment metrics. Our analysis reveals that approximately 15 to 23 percent of attention heads undergo severe disruption during fine-tuning, with lower layers showing greater susceptibility. These findings establish mechanistic foundations for developing targeted mitigation strategies in continual learning systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_18699 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Mechanistic Analysis of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning Imanov, Olaf Yunus Laitinen Machine Learning Computation and Language 68T05, 68T07, 68T50 I.2.6; I.2.7 Large language models exhibit remarkable performance across diverse tasks through pre-training and fine-tuning paradigms. However, continual fine-tuning on sequential tasks induces catastrophic forgetting, where newly acquired knowledge interferes with previously learned capabilities. Despite widespread observations of this phenomenon, the mechanistic understanding remains limited. Here, we present a comprehensive mechanistic analysis of catastrophic forgetting in transformer-based LLMs during sequential fine-tuning. Through systematic experiments across multiple model scales (109B to 400B total parameters) and task sequences, we identify three primary mechanisms driving forgetting: gradient interference in attention weights, representational drift in intermediate layers, and loss landscape flattening. We demonstrate that forgetting severity correlates strongly with task similarity (Pearson r = 0.87) and gradient alignment metrics. Our analysis reveals that approximately 15 to 23 percent of attention heads undergo severe disruption during fine-tuning, with lower layers showing greater susceptibility. These findings establish mechanistic foundations for developing targeted mitigation strategies in continual learning systems. |
| title | Mechanistic Analysis of Catastrophic Forgetting in Large Language Models During Continual Fine-tuning |
| topic | Machine Learning Computation and Language 68T05, 68T07, 68T50 I.2.6; I.2.7 |
| url | https://arxiv.org/abs/2601.18699 |