SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.05992 · ICML · 2026

DSB: Dynamic Sliding Block Scheduling for Diffusion LLMs

Tianwei Zhang, Yonggang Wen, Shenggui Li, Lizhuo Luo

arXiv · PDF · Open in the Atlas

Code that ran

2 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 1 failure of our instrument; 2 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2602.05992.json · badge: /api/ran/2602.05992.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
lizhuo-luo/DSB community
Papers with Code archive
2 ran · 2 no recorded run
FunctionStatusWhere it lives
get_num_transfer_tokens Ran
ran, honoured the contract we drafted
lizhuo-luo/DSB/llada/generate.py
code served (permissive licence) · get_code("e95adfb19ea21d20")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 2, dtype family int, shape (2, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a small block_mask_index tensor
# Shape: (B=2, L=4) with bool dtype
# This represents which positions are masked in the current block
block_mask_index = torch.tensor([
    [True, False, True, False],
    [True, True, False, False]
], dtype=torch.bool)

# Number of steps
steps = 3

ARGS = {
    "block_mask_index": block_mask_index,
    "steps": steps
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family int, shape (2, 3) — shape (2, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_transfer_index_dynamic Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
lizhuo-luo/DSB/llada/generate.py
code served (permissive licence) · get_code("b087fa272381db55")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 3, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Create small tensors for testing
batch_size = 2
seq_len = 3
vocab_size = 4

# logits: shape (batch_size, seq_len, vocab_size)
logits = torch.randn(batch_size, seq_len, vocab_size, dtype=torch.float32)

# temperature: float, can be 0 or positive
temperature = 0.5

# remasking: string, either 'low_confidence' or 'random'
remasking = 'low_confidence'

# mask_index: boolean tensor of shape (batch_size, seq_len)
mask_index = torch.tensor([[True, False, True], [True, True, False]], dtype=torch.bool)

# x: token indices, shape (batch_size, seq_len)
x = torch.randint(0, vocab_size, (batch_size, seq_len), dtype=torch.long)

# num_transfer_tokens: integer or tensor (though it gets overridden in the function)
num_transfer_tokens = 2

# factor: float multiplier for threshold calculation
factor = 1.0

# max_transfer_tokens: optional int limit
max_transfer_tokens = None

ARGS = {
    'logits': logits,
    'temperature': temperature,
    'remasking': remasking,
    'mask_index': mask_index,
    'x': x,
    'num_transfer_tokens': num_transfer_tokens,
    'factor': factor,
    'max_transfer_tokens': max_transfer_tokens,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 3, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
generate_dsb_cache Not yet run
unverified (harvested, no recorded run)
lizhuo-luo/DSB/llada/generate.py
code served (permissive licence) · get_code("77495af1665ed158")
How it ran

no invocation recorded.

get_transfer_index Not yet run
unverified (harvested, no recorded run)
lizhuo-luo/DSB/llada/generate.py
code served (permissive licence) · get_code("a32d174e25e7f535")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 4 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Diffusion large language models (dLLMs) have emerged as a promising alternative for text generation, distinguished by their native support for parallel decoding. In practice, block inference is crucial for avoiding order misalignment in global bidirectional decoding and improving output quality. However, the widely-used fixed, predefined block (naive) schedule is agnostic to semantic difficulty, making it a suboptimal strategy for both quality and efficiency: it can force premature commitments to uncertain positions while delaying easy positions near block boundaries. In this work, we analyze the limitations of naive block scheduling and disclose the importance of dynamically adapting the schedule to semantic difficulty for reliable and efficient inference. Motivated by this, we propose Dynamic Sliding Block (DSB), a training-free block scheduling method that uses a sliding block with a dynamic size to overcome the rigidity of the naive block. To further improve efficiency, we introduce DSB Cache, a training-free KV-cache mechanism tailored to DSB. Extensive experiments across multiple models and benchmarks demonstrate that DSB, together with DSB Cache, consistently improves both generation quality and inference efficiency for dLLMs. Code is released at https://github.com/lizhuo-luo/DSB.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2602.05992")
get_code_for_paper("2602.05992")
have("2602.05992")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2602.05992.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2602.05992.svg)](https://syntology.ai/paper/2602.05992)

Connect an agent — have() is free.