SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2403.12544 · ICLR · 2024

AffineQuant: Affine Transformation Quantization for Large Language Models

Rui Wang, Shilei Wen, Rongrong Ji, Fei Chao, Xiawu Zheng, Yuexiao Ma, Huixia Li, Ling Feng, Xuefeng Xiao

arXiv · PDF · Open in the Atlas

Code that ran

9 of 15 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; 8 failures of our instrument; 6 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; 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/2403.12544.json · badge: /api/ran/2403.12544.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
bytedance/affinequant canonical (classifier label)
classifier basis: census
9 ran · 6 no recorded run
FunctionStatusWhere it lives
ampscaler_get_grad_norm Ran
ran, honoured the contract we drafted
bytedance/affinequant/utils.py
code served (permissive licence) · get_code("4ac3f836e543e81c")
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 Tensor, ndim 0, dtype family float, shape () 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
from math import inf

torch.manual_seed(0)

# Create a small neural network with parameters that have gradients
class SimpleModel(torch.nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = torch.nn.Linear(4, 3)
        self.linear2 = torch.nn.Linear(3, 2)
    
    def forward(self, x):
        return self.linear2(torch.relu(self.linear1(x)))

# Initialize model and compute gradients
model = SimpleModel()
x = torch.randn(2, 4)
y = torch.randn(2, 2)
loss = torch.nn.MSELoss()

output = model(x)
loss_val = loss(output, y)
loss_val.backward()

# Extract parameters with gradients
parameters = [p for p in model.parameters() if p.grad is not None]

# Test case 1: list of parameters with norm_type=2.0
ARGS = {
    "parameters": parameters,
    "norm_type": 2.0
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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 0, dtype family float, shape () — shape () 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
create_logger Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from Nkniexin/SFMP, is pointer only (licence: NONE) · get_code("87877d0c60858bff")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from Nkniexin/SFMP, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

find_layers Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/models/models_utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from utkarshsaxena1/eigenattn, is pointer only (licence: NONE) · get_code("904cd61df2fe8fb9")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from utkarshsaxena1/eigenattn, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

get_act_scales Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/generate_act_scale_shift.py
code served (permissive licence) · get_code("bd1c7aaaa05aeaf3")
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 dict. The image it ran in was not recorded. If your copy returns the same type 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 torch.nn as nn

torch.manual_seed(0)

# Create a simple model with Linear layers
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.linear2(x)
        return x

model = SimpleModel()

# Create a simple dataloader as a list of (x, y) tuples
# Each input tensor has shape (2, 4) - batch_size=2, input_dim=4
dataloader = [
    (torch.randn(2, 4), torch.randint(0, 2, (2,)))
    for _ in range(128)
]

num_samples = 128

ARGS = {
    "model": model,
    "dataloader": dataloader,
    "num_samples": num_samples
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
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_act_shifts Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/generate_act_scale_shift.py
code served (permissive licence) · get_code("550d556e74fb20f7")
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 dict. The image it ran in was not recorded. If your copy returns the same type 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 torch.nn as nn

torch.manual_seed(0)

# Create a small model with Linear layers
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.linear2(x)
        return x

model = SimpleModel()

# Create a small dataloader with (x, y) tuples
# Each sample should have batch size and feature dimension matching model input
dataloader = [
    (torch.randn(2, 4), torch.randn(2, 2)) for _ in range(128)
]

ARGS = {
    "model": model,
    "dataloader": dataloader,
    "num_samples": 128
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
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
make_disjoint_window Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/models/models_utils.py
code served (permissive licence) · get_code("4df9168e0a0c38eb")
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 tuple. The image it ran in was not recorded. If your copy returns the same type 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)

# The function takes a pair (a, b) where both a and b are sequences
# It slices a to remove the last (len(b) - 1) elements
# Both a and b should support len() and slicing operations
# Lists work perfectly for this

a = [1, 2, 3, 4, 5, 6, 7, 8]  # Context window
b = [9, 10, 11]  # Continuation window

pair = (a, b)

ARGS = {
    "pair": pair
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 — a tuple output has no array shape of its own; none is stated
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
sort_layers_by_params Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/parallel_utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from utkarshsaxena1/eigenattn, is pointer only (licence: NONE) · get_code("aaa7723c864e0a5e")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from utkarshsaxena1/eigenattn, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

truncate_number Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/models/transformation.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from saic-fi/mobilequant, is pointer only (licence: NOASSERTION) · get_code("94587954e392b665")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from saic-fi/mobilequant, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

truncate_number Ran
ran; our contract draft was wrong, not the code
bytedance/affinequant/models/models_utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from utkarshsaxena1/eigenattn, is pointer only (licence: NONE) · get_code("4e53fbbdfa160f7d")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from utkarshsaxena1/eigenattn, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

assign_layers_to_gpus Not yet run
unverified (harvested, no recorded run)
bytedance/affinequant/parallel_utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from Aaronhuang-778/SliM-LLM, is pointer only (licence: NONE) · get_code("7e3ef1ed347af0f3")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from Aaronhuang-778/SliM-LLM, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

build_model_and_tokenizer Not yet run
unverified (harvested, no recorded run)
bytedance/affinequant/generate_act_scale_shift.py
code served (permissive licence) · get_code("a041d7c9b413c995")
How it ran

no invocation recorded.

get_lowest_occupied_gpu Not yet run
unverified (harvested, no recorded run)
bytedance/affinequant/parallel_utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from utkarshsaxena1/eigenattn, is pointer only (licence: NONE) · get_code("3c6439a3c61dd2bb")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from utkarshsaxena1/eigenattn, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

get_pile Not yet run
unverified (harvested, no recorded run)
bytedance/affinequant/datautils.py
code served (permissive licence) · get_code("fe6f95071099234c")
How it ran

no invocation recorded.

get_ptb Not yet run
unverified (harvested, no recorded run)
bytedance/affinequant/datautils.py
code served (permissive licence) · get_code("bd4f416e271c2fb5")
How it ran

no invocation recorded.

get_wikitext2 Not yet run
unverified (harvested, no recorded run)
bytedance/affinequant/datautils.py
code served (permissive licence) · get_code("406294899f965c2f")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 14 are also present in other papers whose code we hold (13 with at least one paper dated earlier, 0 the same month, 14 later) and 1 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.

FunctionAlso present inEarliest-datedOthers named
make_disjoint_window 10 other papers
4 dated earlier, 6 later
2304.01089 (2023-04)
RPTQ: Reorder-based Post-training Quantization for Large Language Models
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2310.08041 (2023-10)
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
and 7 more not named here
find_layers 9 other papers
4 dated earlier, 5 later
2304.01089 (2023-04)
RPTQ: Reorder-based Post-training Quantization for Large Language Models
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2310.08041 (2023-10)
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
and 6 more not named here
truncate_number 8 other papers
3 dated earlier, 5 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2310.08041 (2023-10)
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
2402.11295 (2024-02)
OneBit: Towards Extremely Low-bit Large Language Models
and 5 more not named here
truncate_number 8 other papers
2 dated earlier, 6 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2310.08041 (2023-10)
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
2404.14047 (2024-04)
An empirical study of LLaMA3 quantization: from LLMs to MLLMs
and 5 more not named here
create_logger 7 other papers
2 dated earlier, 5 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2401.09417 (2024-01)
Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model
2405.14917 (2024-05)
SliM-LLM: Salience-Driven Mixed-Precision Quantization for Large Language Models
and 4 more not named here
get_lowest_occupied_gpu 6 other papers
3 dated earlier, 3 later
2304.01089 (2023-04)
RPTQ: Reorder-based Post-training Quantization for Large Language Models
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2310.08041 (2023-10)
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
and 3 more not named here
get_ptb 6 other papers
2 dated earlier, 4 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2402.11295 (2024-02)
OneBit: Towards Extremely Low-bit Large Language Models
2404.14047 (2024-04)
An empirical study of LLaMA3 quantization: from LLMs to MLLMs
and 3 more not named here
sort_layers_by_params 6 other papers
3 dated earlier, 3 later
2304.01089 (2023-04)
RPTQ: Reorder-based Post-training Quantization for Large Language Models
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2310.08041 (2023-10)
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
and 3 more not named here
assign_layers_to_gpus 5 other papers
3 dated earlier, 2 later
2304.01089 (2023-04)
RPTQ: Reorder-based Post-training Quantization for Large Language Models
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2310.08041 (2023-10)
QLLM: Accurate and Efficient Low-Bitwidth Quantization for Large Language Models
and 2 more not named here
get_wikitext2 5 other papers
2 dated earlier, 3 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2402.11295 (2024-02)
OneBit: Towards Extremely Low-bit Large Language Models
2404.14047 (2024-04)
An empirical study of LLaMA3 quantization: from LLMs to MLLMs
and 2 more not named here
get_act_scales 3 other papers
1 dated earlier, 2 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2406.01721 (2024-06)
DuQuant: Distributing Outliers via Dual Transformation Makes Stronger Quantized LLMs
2506.03781 (2025-06)
Unifying Uniform and Binary-coding Quantization for Accurate Compression of Large Language Models
get_act_shifts 3 other papers
1 dated earlier, 2 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2406.01721 (2024-06)
DuQuant: Distributing Outliers via Dual Transformation Makes Stronger Quantized LLMs
2506.03781 (2025-06)
Unifying Uniform and Binary-coding Quantization for Accurate Compression of Large Language Models
ampscaler_get_grad_norm 2 other papers
2 later
2407.11062 (2024-07)
EfficientQAT: Efficient Quantization-Aware Training for Large Language Models
2025.acl-long.498 (2025 (year only))
EfficientQAT: Efficient Quantization-Aware Training for Large Language Models
build_model_and_tokenizer 2 other papers
1 dated earlier, 1 later
2308.13137 (2023-08)
OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Models
2506.03781 (2025-06)
Unifying Uniform and Binary-coding Quantization for Accurate Compression of Large Language Models

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

The significant resource requirements associated with Large-scale Language Models (LLMs) have generated considerable interest in the development of techniques aimed at compressing and accelerating neural networks. Among these techniques, Post-Training Quantization (PTQ) has emerged as a subject of considerable interest due to its noteworthy compression efficiency and cost-effectiveness in the context of training. Existing PTQ methods for LLMs limit the optimization scope to scaling transformations between pre-and post-quantization weights. This constraint results in significant errors after quantization, particularly in lowbit configurations. In this paper, we advocate for the direct optimization using equivalent Affine transformations in PTQ (AffineQuant). This approach extends the optimization scope and thus significantly minimizing quantization errors. Additionally, by employing the corresponding inverse matrix, we can ensure equivalence between the pre-and post-quantization outputs of PTQ, thereby maintaining its efficiency and generalization capabilities. To ensure the invertibility of the transformation during optimization, we further introduce a gradual mask optimization method. This method initially focuses on optimizing the diagonal elements and gradually extends to the other elements. Such an approach aligns with the Levy-Desplanques theorem, theoretically ensuring invertibility of the transformation. As a result, significant performance improvements are evident across different LLMs on diverse datasets. Notably, these improvements are most pronounced when using very low-bit quantization, enabling the deployment of large models on edge devices. To illustrate, we attain a C4 perplexity of 15.76 (2.26↓ vs 18.02 in OmniQuant) on the LLaMA2-7B model of W4A4 quantization without overhead. On zero-shot tasks, AffineQuant achieves an average of 58.61% accuracy (1.98% ↑ vs 56.63 in OmniQuant) when using 4/4-bit quantization for LLaMA-30B, which setting a new state-of-the-art benchmark for PTQ in LLMs.

For agents

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

get_harvested_code_for_paper("2403.12544")
get_code_for_paper("2403.12544")
have("2403.12544")

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

curl https://syntology.ai/api/ran/2403.12544.json

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

[![Syntology run record](https://syntology.ai/api/ran/2403.12544.svg)](https://syntology.ai/paper/2403.12544)

Connect an agent — have() is free.