arxiv:2510.02292
Shuyu Wu
wonderwind271
AI & ML interests
LLM (pre)training dynamics; Mechanistic Interpretability
Recent Activity
updated a model 6 days ago
wonderwind271/mib-ioi-4l4h-base published a model 6 days ago
wonderwind271/mib-ioi-4l4h-base updated a model 11 days ago
wonderwind271/mib-ioi-edge-corrOrganizations
models 19
wonderwind271/mib-ioi-4l4h-base
Text Generation • 7.39M • Updated • 472
wonderwind271/mib-ioi-edge-corr
Text Generation • 10.9M • Updated • 211
wonderwind271/mib-ioi-noise001
Text Generation • 10.9M • Updated • 191
wonderwind271/mib-ioi-v0-0714
Text Generation • 10.9M • Updated • 136
wonderwind271/gpt2-probe-finetune-UD-L4
Text Generation • 0.1B • Updated • 3
wonderwind271/childes-checkpoints
Updated
wonderwind271/childes-probing-tunedlens
Updated
wonderwind271/vsdiag-checkpoints
Updated
wonderwind271/gpt2-vsdiag-1
Text Generation • 0.1B • Updated • 9
wonderwind271/gpt2-childes-v4
Text Generation • 0.2B • Updated • 8
datasets 22
wonderwind271/mib-ioi-with-corrupted
Viewer • Updated • 147k • 265
wonderwind271/mib-ioi
Viewer • Updated • 154k • 82
wonderwind271/gsm8k-parsed-expressions
Viewer • Updated • 56k • 19
wonderwind271/speech-dataset
Viewer • Updated • 19.7k • 30 • 1
wonderwind271/PKU-SafeRLHF-transformed
Viewer • Updated • 12k • 7
wonderwind271/en_gum-ud-train
Viewer • Updated • 7.79k • 21
wonderwind271/en_gum-ud-test
Viewer • Updated • 1.19k • 11
wonderwind271/en_gum-ud-dev
Viewer • Updated • 1.23k • 11
wonderwind271/en_ewt-ud-train
Viewer • Updated • 8.04k • 25
wonderwind271/en_ewt-ud-test
Viewer • Updated • 1.09k • 9