The paper claims that framing LLM inference as cognitive load management reduces token consumption without accuracy loss, but the evidence is under-powered and the 'emergent' decomposition is directly trained.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Cognitive Load-Aware Inference: A Neuro-Symbolic Framework for Optimizing the Token Economy of Large Language Models
The paper claims that framing LLM inference as cognitive load management reduces token consumption without accuracy loss, but the evidence is under-powered and the 'emergent' decomposition is directly trained.