R

Ring Lite Linear Preview

Developed by inclusionAI
The Linglong Linear Preview is a hybrid linear sparse large language model open-sourced by InclusionAI, with a total of 17.1B parameters and 3.0B activated parameters. This model implements long-text reasoning based on a hybrid linear attention mechanism, achieving near-linear computational complexity and near-constant space complexity during inference.
Downloads 25
Release Time : 4/24/2025

Model Overview

The Linglong Linear Preview is a large language model based on a hybrid linear attention mechanism, significantly reducing computational overhead during training and inference while delivering high performance in long-text reasoning scenarios.

Model Features

Efficient Long-Text Reasoning
Achieves near-linear computational complexity and near-constant space complexity through a hybrid linear attention mechanism, significantly improving long-text reasoning efficiency.
High-Performance Inference
Scores 55.0 on the AIME24 test and 93.8 on the MATH-500 test, performing comparably to DeepSeek-R1-Distill-Qwen-7B.
High Throughput
In vLLM tests, throughput exceeds twice that of softmax attention models of the same scale.

Model Capabilities

Long-Text Generation
Chinese Text Understanding
English Text Understanding
Mathematical Reasoning

Use Cases

Education
Math Problem Solving
Solving complex math problems
Achieved 93.8 on the MATH-500 test.
Research
Long-Text Analysis
Processing and analyzing long-text content
Supports 64K context length.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase