Modeling LLM dialogues as bridging-inference knowledge graphs reveals more stable and coherent personas than traditional lexical or stylistic analysis methods.
In: Proceedings of the 2024 Conference of the North American Chapter of the As- sociation for Computational Linguistics: Human Language Technologies (NAACL- HLT)
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
-
The Pragmatic Persona: Discovering LLM Persona through Bridging Inference
Modeling LLM dialogues as bridging-inference knowledge graphs reveals more stable and coherent personas than traditional lexical or stylistic analysis methods.