Lunar Society LabsRESEARCH LIBRARY
⌘ K 中文
THE RESEARCH ATLAS

DeepSeek Papers & Recent Research

Institution base · DeepSeek 🇨🇳 China

Reasoning · MoE · Efficient training. Selected primary papers, reports, model cards and official technical articles. Checked 2026-10-25. Product announcements are not presented as papers.

Official research source ↗GitHub ↗
2026-09-10 · DeepSeek · Official research articleInstitution base · DeepSeek 🇨🇳 China

DeepSeek-V4.1-Flash: smarter, faster, more efficient

Official release introducing an asymmetric causal encoder-decoder architecture and linking its technical report and model resources.

2025-12 · DeepSeek · Technical reportInstitution base · DeepSeek 🇨🇳 China

DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models

Sparse attention and post-training methods are combined to improve efficiency, reasoning and agent-oriented behavior.

2025-01 · DeepSeek · Research paperInstitution base · DeepSeek 🇨🇳 China

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

Reinforcement learning is used to develop reasoning behavior, with additional pipelines for usable models and distillation.

2024-12 · DeepSeek · Technical reportInstitution base · DeepSeek 🇨🇳 China

DeepSeek-V3 Technical Report

A mixture-of-experts language model combines architectural and training-efficiency techniques, including latent attention and multi-token prediction.

What this index holds from DeepSeek

This index currently holds 4 records attributed to DeepSeek (2024–2026). Each entry links to the primary source; product announcements and news posts are not counted as papers.

Related work on the same topics

These records come from other institutions but address the same topics, so the lab can be read in context.