YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

cayley-10b-k8-2l-mlp_in

205M-param GPT with 2-level CayleySAE at mlp_in. Trained on FineWeb-Edu-10B for 16k iters (~10.5B tokens). Reaches 3.228 CE val loss.

CayleySAE config

Level n k delta Features
L0 10 8 0 (root) 1,024
L1 13 16 64 8,192
  • Total features per layer: 9,216
  • Active per token: 24
  • per_parent_budget=True, score_standardize=True
  • Location: mlp_in

Training config

optimizer: muon (lr=0.006), adamw (lr=0.006)
lr_schedule: linear_warmdown, warmdown_frac=0.2
batch_size: 40 × 2 GPUs, gradient_accumulation_steps: 8 (effective batch: 640)
seq_len: 1024, max_iters: 16000
Downloads last month
10
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support