Back to feed
arXiv cs.CL
arXiv cs.CL
7/22/2026
Computational models of pragmatic reasoning with flexible generation of meaning and expression alternatives

Computational models of pragmatic reasoning with flexible generation of meaning and expression alternatives

Short summary

Researchers propose SAGE, a framework combining cognitive models with language models for pragmatic reasoning. SAGE uses LM-based proposers to generate alternatives, evaluators to assess them, and selectors for rule-based computational steps. Across three case studies, LM proposers performed well but LM evaluators struggled with formal judgments, revealing both promise and limitations of neuro-symbolic pragmatic models.

  • •SAGE combines cognitive models with LMs for pragmatic language reasoning using proposers, evaluators, and selectors
  • •LM proposers reliably generate alternatives; LM evaluators better at intuitive than formal judgments
  • •Tested on referential expression generation, M-implicatures, and Gricean implicatures with human data comparisons

Generated with AI, which can make mistakes.

#research-breakthrough#ai-tools
Read full article at arXiv cs.CL

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

Evaluating SageMath-Augmented LLM Agents for Computational and Experimental Mathematics
arXiv CS.AI
Enhancing Small Language Models Reasoning through Knowledge Graph Grounding
arXiv CS.AI
The user wants me to rewrite a headline for a mobile feed based on the original title and summary.
arXiv CS.AI
Developmental Trajectories of Situation Modeling and Mentalizing in Transformer Language Models
arXiv cs.CL