Models trained as a part of the "Attention Drift: What Speculative Decoding Models Learn" paper, shared for reproducing experiments.
Doğaç Eldenk
Dogacel
AI & ML interests
Inference Acceleration
Recent Activity
authored a paper 30 days ago
Attention Drift: What Autoregressive Speculative Decoding Models Learn updated a collection about 1 month ago
Attention Drift upvoted a paper about 1 month ago
Attention Drift: What Autoregressive Speculative Decoding Models LearnOrganizations
None yet
models 21
Dogacel/specdrift-gpt-oss-20b-eagle3
0.4B • Updated • 40 • 1
Dogacel/specdrift-gpt-oss-120b-eagle3
0.4B • Updated • 34
Dogacel/specdrift-qwen3.6-35b-a3b-eagle3
0.2B • Updated • 109 • 3
Dogacel/specdrift-qwen3.6-27b-eagle3
0.6B • Updated • 55 • 1
Dogacel/paper-attention-drift-qwen35-9b-post
0.4B • Updated • 3
Dogacel/paper-attention-drift-qwen35-9b-pre
0.4B • Updated • 3
Dogacel/paper-attention-drift-gpt-20b-pre
0.3B • Updated • 3
Dogacel/paper-attention-drift-gpt-20b-post
0.3B • Updated • 3
Dogacel/paper-attention-drift-qwen3-pre
0.4B • Updated • 2
Dogacel/paper-attention-drift-qwen3-post
0.4B • Updated • 5
datasets 4
Dogacel/nemotron-post-training-v2-gpt-oss-20b-regen
Viewer • Updated • 931k • 19
Dogacel/nemotron-post-training-v2-gpt-oss-120b-regen
Viewer • Updated • 930k • 46
Dogacel/nemotron-post-training-v2-qwen-3.5-0.8b-regen
Viewer • Updated • 932k • 159
Dogacel/nemotron-post-training-v2-qwen-3.5-9b-regen
Viewer • Updated • 940k • 35