-
sDPO: Don't Use Your Data All at Once
Paper • 2403.19270 • Published • 41 -
Advancing LLM Reasoning Generalists with Preference Trees
Paper • 2404.02078 • Published • 46 -
Learn Your Reference Model for Real Good Alignment
Paper • 2404.09656 • Published • 90 -
mDPO: Conditional Preference Optimization for Multimodal Large Language Models
Paper • 2406.11839 • Published • 40
zyyang
zy0yang
·
AI & ML interests
SFT & RLHF
Organizations
Alignment-DPO-line
-
sDPO: Don't Use Your Data All at Once
Paper • 2403.19270 • Published • 41 -
Advancing LLM Reasoning Generalists with Preference Trees
Paper • 2404.02078 • Published • 46 -
Learn Your Reference Model for Real Good Alignment
Paper • 2404.09656 • Published • 90 -
mDPO: Conditional Preference Optimization for Multimodal Large Language Models
Paper • 2406.11839 • Published • 40
MOE
models
0
None public yet
datasets
0
None public yet