# Common sense reasoning

MobileLLM is an optimized Transformer architecture language model developed by Meta, specifically designed for resource-constrained edge-side applications, with parameter scales ranging from 125M to 1.5B.

MobileLLM is Meta's series of sub-billion parameter language models specifically optimized for resource-constrained devices, significantly improving on-device inference efficiency through deep-narrow architecture design.

A generative model supporting Russian casual conversation and common sense reasoning, based on GPT-2 architecture

A GPT-2 fine-tuned model for common sense verification and explanation, used to generate reasons for statements that violate common sense