Subgraph Atlas · centered on entity

Rejection Sampling Fine-Tuning

technique0 mentions· velocity: stable

Sampling multiple completions, scoring with a reward model, and fine-tuning on the top samples — a simpler alternative to PPO used in Llama 2.

Two-hop subgraph: this entity, every entity it directly relates to, and every entity those neighbors relate to. Drag a node, scroll to zoom, click to inspect — or click any neighbor and re-center the atlas there.

0 nodes · 0 edges · loading…

companypersonai_modelproductresearch_labbenchmarkframework

drag to move · scroll to zoom · click a node

Top connections

Meta AIproduct

5 mentions

→ Center atlas here

Reinforcement Learning from Human Feedback (RLHF)technique

4 mentions

→ Center atlas here

Llama 2: Open Foundation and Fine-Tuned Chat Modelspaper

0 mentions

→ Center atlas here

How to read this: the white-ringed node is Rejection Sampling Fine-Tuning. Surrounding nodes are direct relationships; the second ring is what those neighbors connect to. Edge thickness scales with source-article evidence. Click any node and choose Center graph here to walk the graph.