LDLMIn construction

All models, with their blocks →

Size
306M trained
Published
Meshchaninov et al. · May 2026
Line
Diffusion language model
View
Architecture
Source

LDLM @ e18c42a, not yet public · arXiv:2605.07933v1 · OpenWebText · 1024 · checkpoint, tokenizer and saved statistics not verifiedHow to read

TrainingGeneration

Token IDs

×

tensor

Integer vocabulary indices, one for each position, handed to the detokenizer.

shape
II [B batch, T positions]
Sources

Contextual evidence from Choose tokens

Generation: the sampled latents pass the latent decoder and the token decoder, each position takes the token of the largest logit, and the tokens are decoded to text, special tokens kept unless truncation is on.

The implementation, pinned at revision e18c42acdfb7, not yet public

Contextual evidence from Detokenizer

Generation: sequences are not truncated at the end token, and the trained adaptive time grid is used.

The implementation, pinned at revision e18c42acdfb7, not yet public

Contextual evidence from Detokenizer

The tokenizer is loaded by name, set up, and gains a `<mask>` token when it has none; the vocabulary size is its length.

The implementation, pinned at revision e18c42acdfb7, not yet public