LLM retrieval accuracy decays log-linearly with the number of interfering prior updates, and 'forget' instructions barely improve it.
In-context learning and induction heads, March 2022
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
extension 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
extension 1polarities
extend 1representative citing papers
citing papers explorer
-
Unable to Forget: Proactive Interference Reveals Working Memory Limits in LLMs Beyond Context Length
LLM retrieval accuracy decays log-linearly with the number of interfering prior updates, and 'forget' instructions barely improve it.