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_gr65 mode: global_recurrence66 baseline_kwargs:67 ffn_w: 64068model:69 name: tiny_p4v270 vocab_size: 12871 d_model: 25672 n_layers: 673 n_heads: 874 n_kv_heads: 475 ffn_width: 76876 max_seq_len: 32077 rope_theta: 10000.078 norm_eps: 1.0e-0579 tie_embeddings: true80 init_std: 0.0281 use_workspace: true82 w: 6483 ffn_w: 19284 n_slots: 1685 n_editor_blocks: 286 n_editor_heads: 287 r_max: 888 eta_max: 0.2589 clip_c: 1.090 clip_eps: 1.0e-0691 readout_zero_init: false92 readout_init_std: 0.293 readout_scale: 1.094 controller_hidden: 095 ctrl_input: z96 ctrl_use_entropy: false97 lambda_cost: 0.098 kappa_risk: 0.099 dense_writes: false100resolved_train:101 batch_size: 32102 seq_len: 320103 steps: 3000104 lr: 0.001105 warmup: 100106 weight_decay: 0.1107 grad_clip: 1.0108 log_every: 50109 eval_every: 500110 eval_batches: 12111 ckpt_every: 3000112