QSTRBench is a new benchmark evaluating LLMs on compositional reasoning, converse relations, and conceptual neighbourhoods across QSTR calculi including a newly published RCC-22 CN, showing models exceed chance but fail to achieve consistent correctness.
Extending Binary Qualitative Direction Calculi with a Granular Distance Concept: Hidden Feature Attachment
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
In this paper we introduce a method for extending binary qualitative direction calculi with adjustable granularity like OPRAm or the star calculus with a granular distance concept. This method is similar to the concept of extending points with an internal reference direction to get oriented points which are the basic entities in the OPRAm calculus. Even if the spatial objects are from a geometrical point of view infinitesimal small points locally available reference measures are attached. In the case of OPRAm, a reference direction is attached. The same principle works also with local reference distances which are called elevations. The principle of attaching references features to a point is called hidden feature attachment.
fields
cs.AI 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
QSTRBench: a New Benchmark to Evaluate the Ability of Language Models to Reason with Qualitative Spatial and Temporal Calculi
QSTRBench is a new benchmark evaluating LLMs on compositional reasoning, converse relations, and conceptual neighbourhoods across QSTR calculi including a newly published RCC-22 CN, showing models exceed chance but fail to achieve consistent correctness.