Entity · technique

referential grounding

techniqueactivereferential-grounding-e55c5d86·1 events·first seen Jun 1, 2026

Aliases: referential grounding

Co-occurring entities

large language models compositional generalization Personal Relation Task Paperno 2022

More like this (12)

vision-language grounding Deceptive Grounding: Entity Attribution Failure in Clinical Retrieval-Augmented Generation deceptive grounding GUI grounding Grounding latent algorithm routing in transformer reasoning Grounding LLM Reasoning under Incomplete Graph Evidence reference attention Elastic Semantic Anchoring Decoupled Search Grounding Analogical Deep Research Semantic Reference Frames Grounded Context Preference Dataset

Recent events (1)

5arXiv · cs.CL·Jun 1, 2026·source ↗

LLMs Show Inverted Compositional Strengths vs. Humans on Reference Resolution Task

This paper evaluates LLMs and humans on the Personal Relation Task (Paperno 2022), distinguishing between Extensional tasks (resolving what an expression refers to) and Intensional tasks (representing structured sense/formula). The study finds that humans outperform LLMs on Extensional tasks while LLMs outperform humans on Intensional tasks—an inverted pattern of strengths. The authors argue this asymmetry reflects the absence of referential grounding in LLM training as a key gap in human-like language understanding.

Evaluation and Benchmarking Alignment and RLHF large language models referential grounding compositional generalization +2 more