1experiment:2 seed: 13 model:4 name: tiny_p4v25 vocab_size: 1286 d_model: 2567 n_layers: 68 n_heads: 89 n_kv_heads: 410 ffn_width: 76811 max_seq_len: 32012 w: 6413 ffn_w: 19214 n_slots: 1615 n_editor_heads: 216 r_max: 817 readout_zero_init: false18 readout_init_std: 0.219 init:20 backbone_checkpoint: experiments/runs/p4_tiny_backbone_v2_seed01_20261029233624/checkpoint21 freeze_backbone: true22 auto_readout_scale: true23 data:24 kind: synthetic25 tokenizer: char26 splits_dir: data/splits/synthetic_v127 loss_span: answer28 eval_splits:29 - dev_iid30 - dev_template31 - dev_depth32 - dev_compose33 budgets:34 - 035 - 136 - 237 - 438 - 839 budget_probs:40 - 0.241 - 0.242 - 0.243 - 0.244 - 0.245 budget_per: position46 train:47 batch_size: 3248 seq_len: 32049 steps: 300050 lr: 0.00151 warmup: 10052 weight_decay: 0.153 grad_clip: 1.054 log_every: 5055 eval_every: 50056 eval_batches: 1257 ckpt_every: 300058 em_budgets:59 - 060 - 161 - 262 - 863 em_max_items: 100064 name: p4v2_dense65 mode: dense66model:67 name: tiny_p4v268 vocab_size: 12869 d_model: 25670 n_layers: 671 n_heads: 872 n_kv_heads: 473 ffn_width: 76874 max_seq_len: 32075 rope_theta: 10000.076 norm_eps: 1.0e-0577 tie_embeddings: true78 init_std: 0.0279 use_workspace: true80 w: 6481 ffn_w: 19282 n_slots: 1683 n_editor_blocks: 284 n_editor_heads: 285 r_max: 886 eta_max: 0.2587 clip_c: 1.088 clip_eps: 1.0e-0689 readout_zero_init: false90 readout_init_std: 0.291 readout_scale: 1.092 controller_hidden: 093 ctrl_input: z94 ctrl_use_entropy: false95 lambda_cost: 0.096 kappa_risk: 0.097 dense_writes: true98resolved_train:99 batch_size: 32100 seq_len: 320101 steps: 3000102 lr: 0.001103 warmup: 100104 weight_decay: 0.1105 grad_clip: 1.0106 log_every: 50107 eval_every: 500108 eval_batches: 12109 ckpt_every: 3000110