Language-model training and adaptation · MLD-1047
RLVR
Reinforcement learning using rewards from checkable outcomes such as exact answers or executable tests.
Original reference page · detailed expansion pending. Original definitions, examples, notes and citations are preserved below. New examples are illustrative, not benchmark results.
Original example
Reward code that passes valid unit tests.
Original caveat
A verifiable proxy can still be incomplete or exploitable.
Context and aliases
Level: Intermediate. Stages: Training · Adaptation · Evaluation.
Tags: DL · SLM · LLM · NLP.
Aliases: Reinforcement learning with verifiable rewards.
Related concepts
Concepts linking here
Original references
- Transformers glossary ↗ · Consulted
- PEFT: LoRA conceptual guide ↗ · Consulted
- TRL: supervised fine-tuning ↗ · Reference link
- TRL documentation ↗ · Consulted
- Original reference URL ↗