DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Fields

id

sources/hf-eco-deepseek-r1-paper

URL

https://arxiv.org/abs/2501.12948

Publisher

DeepSeek-AI

Kind

paper

Published

2025-01-22

Accessed

2026-10-03

notes

Referenced by every DeepSeek-R1-Distill model card (arxiv:2501.12948 tag). Source for the claim that the distill weights are a reasoning-distilled variant of an existing dense architecture, not a new architecture.

Notes

Referenced by every DeepSeek-R1-Distill model card (arxiv:2501.12948 tag). Source for the claim that the distill weights are a reasoning-distilled variant of an existing dense architecture, not a new architecture.

Cross-references

This record references no other records.