nics-efc
/

Standard-1.7B

Text Generation

looped transformer

text-generation-inference

Model card Files Files and versions

This is the general version of Standard-1.7B, trained on a mixture of math, code, and science data, presented in the paper Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models.

Please visit our GitHub repo for more information.

Sample Usage

Please see Github Example for sample usage.

Downloads last month: 8

Safetensors

Model size

2B params

Tensor type

BF16

·

Model tree for nics-efc/Standard-1.7B

Base model

Qwen/Qwen3-1.7B-Base

Finetuned

(339)

this model

Quantizations

Dataset used to train nics-efc/Standard-1.7B

Collection including nics-efc/Standard-1.7B

TaH

Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models • 4 items • Updated Nov 23, 2025 • 2

Paper for nics-efc/Standard-1.7B

Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models

Paper • 2511.08577 • Published Nov 11, 2025 • 106