Understanding LLM Performance Degradation in Multi-Instance Processing: The Roles of Instance Count and Context Length.

Jingxuan Chen, Mohammad Taher Pilehvar, Jos Camacho-Collados

Browse the full ACL paper archive.