CoPE is a drop-in enhancement of RoPE that delivers consistent gains within the training context and during long-context extrapoaltion.
Haoran Li PRO
haoranli-ml
·
AI & ML interests
ML, RL, Foundation Models
Recent Activity
upvoted a paper 3 days ago
RRSI: Regularized Recursive Self-Improvement of Agent Harnesses updated a dataset 4 months ago
haoranli-ml/dapo-2k published a dataset 4 months ago
haoranli-ml/dapo-2kOrganizations
models 48
haoranli-ml/NOBOS-lcft_gemma-2b_prolong-gemma-parts_ProLong64KMix_bsz256_steps1250_lr1e-5_warmup0.1_cope
Text Generation • 3B • Updated • 22
haoranli-ml/genvf-v8-tcs-4B-v2
Updated
haoranli-ml/genvf-v8-tcs-4B
Updated
haoranli-ml/genvf-v8-8B
Updated
haoranli-ml/genvf-v9_1-nextN-proof-only-8B
Updated
haoranli-ml/genvf-v9_1-nextN-start-300
Updated
haoranli-ml/genvf-v9_1-nextN
Updated
haoranli-ml/genvf-v9_1-nextN-proof-only
Updated
haoranli-ml/genvf-v9-nextN-proof-only
Updated
haoranli-ml/genvf-v9-nextN
Updated
datasets 151
haoranli-ml/dapo-2k
Viewer • Updated • 2k • 10
haoranli-ml/IMOProofBench-Adv
Viewer • Updated • 30 • 11
haoranli-ml/prolong-ultrachat-gemma
Updated • 67
haoranli-ml/prolong-data-64K-gemma
Updated • 7.14k
haoranli-ml/genvf-data-generator-100prefix-v1
Viewer • Updated • 110 • 10
haoranli-ml/genvf-filtered-proof-graded_score7_only_with_summaries-full
Viewer • Updated • 1.11k • 10
haoranli-ml/genvf-tcs-full-prl-train-prompt_v2
Viewer • Updated • 2.78k • 4
haoranli-ml/genvf-tcs-full-prl-train
Viewer • Updated • 2.78k • 8
haoranli-ml/genvf-tcs-full
Viewer • Updated • 1.42k • 14
haoranli-ml/tcs-stage1-suffixes-cross
Viewer • Updated • 2.32k • 7