diff options
| -rw-r--r-- | README.md | 8 |
1 files changed, 6 insertions, 2 deletions
@@ -1,4 +1,8 @@ -# GRAFT (KAFT): Topology-Factorized Jacobian-Aligned Feedback for Deep GNNs +# KAFT: Topology-Factorized Jacobian-Aligned Feedback for Deep GNNs + +> Internal class names (`GraphGrAPETrainer`, file paths starting with +> `graft_*` / dataset keys `'GRAFT'`) are the original code identifier. +> The method is referred to as **KAFT** in the paper. Code release accompanying the NeurIPS 2026 submission. @@ -20,7 +24,7 @@ computed in O(1) parallel depth on GPUs. ``` src/ core method - trainers.py BPTrainer, GraphGrAPETrainer (KAFT), DFA/DFA-GNN, alignment + trainers.py BPTrainer, GraphGrAPETrainer (= KAFT), DFA/DFA-GNN, alignment data.py PyG dataset loaders, normalized  / row-Â, sparse-mm helpers experiments/ one runner per reported result block (see `## Reproducing`) figures/ figure generators + the four rendered PDFs in the paper |