Sparse Sparse Attention by Aligning Full and Sparse Attention Outputs in Feature Space
Junrulu
Junrulu
AI & ML interests
None yet
Recent Activity
updated
a collection
15 days ago
SSA
updated
a collection
15 days ago
SSA
updated
a collection
15 days ago
SSA
Organizations
Eliminating Biased Length Reliance of Direct Preference Optimization via Down-Sampled KL Divergence
-
jiazhengli/Pythia-2.8B-HH-RLHF-Iterative-SamPO
Text Generation • 3B • Updated -
jiazhengli/Pythia-2.8B-TLDR-Iterative-SamPO
Text Generation • 3B • Updated • 2 -
Junrulu/Llama-3-8B-Instruct-Iterative-SamPO
Text Generation • 8B • Updated • 10 • 1 -
Eliminating Biased Length Reliance of Direct Preference Optimization via Down-Sampled KL Divergence
Paper • 2406.10957 • Published • 2
Sparse Sparse Attention by Aligning Full and Sparse Attention Outputs in Feature Space
A Fine-Grained Composite Benchmark for Role-Playing and Instruction-Following
Eliminating Biased Length Reliance of Direct Preference Optimization via Down-Sampled KL Divergence
-
jiazhengli/Pythia-2.8B-HH-RLHF-Iterative-SamPO
Text Generation • 3B • Updated -
jiazhengli/Pythia-2.8B-TLDR-Iterative-SamPO
Text Generation • 3B • Updated • 2 -
Junrulu/Llama-3-8B-Instruct-Iterative-SamPO
Text Generation • 8B • Updated • 10 • 1 -
Eliminating Biased Length Reliance of Direct Preference Optimization via Down-Sampled KL Divergence
Paper • 2406.10957 • Published • 2
Tuning LLMs to Use Memos for Consistent Long-Range Open-Domain Conversation