learn.ifs.name

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

DeepSeek-AI, 2025, DeepSeek-R1 (arXiv preprint arXiv:2501.12948)

Subscriber deck — free during early access

1 cards in “2.3.4 Reinforcement Learning for all Scenarios” · whole deck

2.3.4 Reinforcement Learning for all Scenarios

  1. What did DeepSeek-R1’s final RL stage add?
    Helpfulness and harmlessness alongside reasoning: rule-based rewards for math, code and logic, and reward models of human preferences for general prompts. Helpfulness was judged on the final summary only, harmlessness on the whole response including the reasoning.
    • DeepSeek-AI, 2025, DeepSeek-R1 (arXiv preprint arXiv:2501.12948), §2.3.4

Source

  • DeepSeek-AI. 2025. DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning. In arXiv preprint arXiv:2501.12948. https://arxiv.org/abs/2501.12948