1experiment:2 name: p4_editor_v13 seed: 14 mode: prisme5 model:6 name: tiny_p47 vocab_size: 1288 d_model: 1929 n_layers: 410 n_heads: 611 n_kv_heads: 212 ffn_width: 51213 max_seq_len: 32014 w: 6415 ffn_w: 19216 n_slots: 1617 n_editor_heads: 218 r_max: 819 init:20 backbone_checkpoint: experiments/runs/p4_tiny_backbone_seed01_20261029232626/checkpoint21 freeze_backbone: true22 data:23 kind: synthetic24 tokenizer: char25 splits_dir: data/splits/synthetic_v126 loss_span: answer27 eval_splits:28 - dev_iid29 - dev_template30 - dev_depth31 - dev_compose32 budgets:33 - 034 - 135 - 236 - 437 - 838 budget_probs:39 - 0.240 - 0.241 - 0.242 - 0.243 - 0.244 budget_per: position45 train:46 batch_size: 3247 seq_len: 32048 steps: 300049 lr: 0.00150 warmup: 10051 weight_decay: 0.152 grad_clip: 1.053 log_every: 5054 eval_every: 50055 eval_batches: 856 ckpt_every: 150057 em_budgets:58 - 059 - 260 - 861 em_max_items: 40062model:63 name: tiny_p464 vocab_size: 12865 d_model: 19266 n_layers: 467 n_heads: 668 n_kv_heads: 269 ffn_width: 51270 max_seq_len: 32071 rope_theta: 10000.072 norm_eps: 1.0e-0573 tie_embeddings: true74 init_std: 0.0275 use_workspace: true76 w: 6477 ffn_w: 19278 n_slots: 1679 n_editor_blocks: 280 n_editor_heads: 281 r_max: 882 eta_max: 0.2583 clip_c: 1.084 clip_eps: 1.0e-0685 readout_zero_init: true86 controller_hidden: 087 lambda_cost: 0.088 kappa_risk: 0.089 dense_writes: false90resolved_train:91 batch_size: 3292 seq_len: 32093 steps: 300094 lr: 0.00195 warmup: 10096 weight_decay: 0.197 grad_clip: 1.098 log_every: 5099 eval_every: 500100 eval_batches: 8101 ckpt_every: 1500102