guided

Definition

Marks generation or search steered by an external signal rather than left to the model alone. The canonical case is classifier-free guidance in diffusion, where a single scale trades sample fidelity against diversity by amplifying the conditional direction. The same word covers guided decoding, where a grammar, a verifier, or retrieved evidence constrains the token stream. What unites them is intervening during generation rather than filtering afterwards.

Explainers using this term

• Paper Explained: RewardVerse — A Video Reward Model That Writes the Rubric Before It Scores (RGPO)RewardVerse: Rubric-Guided Policy Optimization for Video Reward Modeling
• Paper walkthrough: BeaconKV — evicting KV cache by predicting when a reasoning model will re-read its own planBeaconKV: Key-Value Cache Compression Guided by Beacon Queries for Efficient Large Reasoning Model Inference
• Paper Walkthrough: ENEAS — Telling Real Instances From Their DoppelgangersENEAS: Embedding-guided Neural Ensemble for Adaptive Segmentation