CluE, a cluster-based self-evolving prompt strategy, achieves a 9.04% relative gain on the new BEHEMOTH benchmark for heterogeneous LLM memory extraction and outperforms prior frameworks.
Figure 14: Prompt evolved by CluE fromSimple, using Qwen3-32B
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Self-Evolving LLM Memory Extraction Across Heterogeneous Tasks
CluE, a cluster-based self-evolving prompt strategy, achieves a 9.04% relative gain on the new BEHEMOTH benchmark for heterogeneous LLM memory extraction and outperforms prior frameworks.