Knowledge Graph

Graph explorer

Search resolved entities, walk their edges, and read the sentence behind every one.

Entities25 of 207,987

Entity

upside-down reinforcement learning

technology1 mentionslast seen Oct 16, 2026

A reinforcement learning approach combined with synthetic data distillation to train models for prompt generation.

Relationships 1

small language modeldepends onupside-down reinforcement learning88% confidence1 source

upside-down reinforcement learning

Relationship

small language modeldepends onupside-down reinforcement learning

88% confidence1 supporting mentions

Either name opens that entity’s relationships.

Sources 1

preprintarXivDec 3, 2025Scaling Multimodal Search and Recommendation with Small Language Models via Upside-Down Reinforcement Learning
combines upside-down reinforcement learning with synthetic data distillation from a large language model (Llama-3) to train a 100M-parameter GPT-2 model