Large Language Models
25 views
DPO
Quick Definition
Alignment method bypassing reward model for preference learning
Full Definition
Direct Preference Optimization, a simplified alignment method that bypasses explicit reward model training.
Examples
LLM alignment, preference learning, simpler RLHF alternative
Related Terms
rlhf
alignment
reward-model
More Large Language Models Terms
Training LLM
Adjusting parameters on large datasets to learn language
Function Calling
Structured mechanism for LLMs to generate API function calls
Nucleus Sampling
Generation using cumulative probability threshold for token selection
Inference LLM
Using a trained LLM to generate outputs from inputs
Grounding LLM
Connecting LLM outputs to verifiable external sources
Mixture of Experts
Architecture routing tokens to specialized expert sub-networks