LLMs display high variance and major accuracy drops on GSM-Symbolic variants of grade-school math problems, indicating they replicate training patterns rather than execute logical reasoning.
The algebraic mind: Integrating connectionism and cognitive science
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
LLMs represent semantic relations geometrically via embedding distance and direction; a linear Polar Probe decodes these structures from middle-layer activations and generalizes to new entities.
citing papers explorer
-
GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models
LLMs display high variance and major accuracy drops on GSM-Symbolic variants of grade-school math problems, indicating they replicate training patterns rather than execute logical reasoning.
-
Polar probe linearly decodes semantic structures from LLMs
LLMs represent semantic relations geometrically via embedding distance and direction; a linear Polar Probe decodes these structures from middle-layer activations and generalizes to new entities.