{
 "input_dim": 512,
 "hidden_dim": 2048,
 "output_dim": 2048,
 "memory_dim": 512,
 "memory_slots": 64,
 "num_layers": 12,
 "num_associations": 8,
 "use_quantization": false,
 "use_mla": false,
 "mla_latent_dim": 32,
 "use_causal_lm": true,
 "vocab_size": 5000,
 "max_seq_len": 128,
 "causal_window_size": 64,
 "sae_k": 64,
 "ntm_memory_slots": 16,
 "d_model": 512,
 "mla_n_heads": 8,
 "mla_max_cache_len": 4096,
 "lm_num_attn_layers": 2,
 "lm_pooling": "mean"
}