SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2404.13844 · 2024

ColA: Collaborative Adaptation with Gradient Learning

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 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; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 2 failures of our instrument; 0 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/2404.13844.json · badge: /api/ran/2404.13844.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
diaoenmao/cola-collaborative-adaptation-with-gradient-learning canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
FunctionStatusWhere it lives
make_all_controls Ran
ran; our contract draft was wrong, not the code
diaoenmao/cola-collaborative-adaptation-with-gradient-learning/src/process.py
code served (permissive licence) · get_code("ce25fb1505eb2e02")
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 list, ndim 2, 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 itertools

num_experiments = 3

exp = [str(x) for x in list(range(num_experiments))]

def make_controls(control_name):
    control_names = []
    for i in range(len(control_name)):
        control_names.extend(list(('_'.join(x) for x in itertools.product(*control_name[i]))))
    controls = [exp] + [control_names]
    controls = list(itertools.product(*controls))
    return controls

def make_all_controls(mode, task_name):
    if task_name == 's2s':
        data_names = ['fpb-sa', 'wikisql', 'samsum', 'e2enlg', 'webnlg-2017', 'dart']
        model_names = ['bart-base']
    elif task_name == 'clm':
        data_names = ['dolly-15k']
        model_names = ['gpt2']
    elif task_name == 'sc':
        data_names = ['glue-cola', 'glue-mnli', 'glue-mrpc', 'glue-qnli', 'glue-qqp', 'glue-rte', 'glue-sst2', 'glue-stsb']
        model_names = ['roberta-base']
    elif task_name == 'ic':
        data_names = ['MNIST', 'CIFAR10']
        model_names = ['linear', 'mlp', 'cnn']
    else:
        raise ValueError('Not valid task name')
    if mode == 'full':
        if task_name == 'ic':
            batch_size = ['256']
        else:
            batch_size = ['32']
        control_name = [[data_names, model_names, [task_name], ['full'], batch_size]]
        controls = make_controls(control_name)
    elif mode == 'peft':
        if task_name == 'ic':
            ft_name = ['lora']
            batch_size = ['256']
        else:
            ft_name = ['lora', 'adalora', 'ia3', 'promptune', 'prefixtune', 'ptune']
            if model_names[0] == 'llama-2':
                batch_size = ['8']
            else:
                batch_size = ['32']
        control_name = [[data_names, model_names, [task_name], ft_name, batch_size]]
        controls = make_controls(control_name)
    elif mode == 'cola':
        ft_name = ['cola-lowrank-1', 'cola-linear-1', 'cola-mlp-1']
        if task_name == 'ic':
            batch_size = ['256']
        elif model_names[0] == 'llama-2':
            batch_size = ['8']
        else:
            batch_size = ['32']
        control_name = [[data_names, model_names, [task_name], ft_name, batch_size]]
        controls = make_controls(control_name)
    elif mode == 'cola_step':
        ft_name = ['cola-lowrank-1', 'cola-lowrank-2', 'cola-lowrank-4', 'cola-lowrank-8']
        if task_name == 'ic':
            batch_size = ['64']
        else:
            batch_size = ['8']
        control_name = [[data_names, model_names, [task_name], ft_name, batch_size]]
        controls = make_controls(control_name)
    elif mode == 'cola_dist':
        data_names = ['dolly-15k']
        ft_name = ['cola-lowrank-1', 'cola-lowrank~linear-1', 'cola-lowrank~mlp-1']
        if model_names[0] == 'llama-2':
            batch_size = ['8']
        else:
            batch_size = ['32']
        dist_mode = ['alone']
        control_name = [[data_names, model_names, [task_name], ft_name, batch_size, dist_mode]]
        controls = make_controls(control_name)
    elif mode == 'cola_merge':
        ft_name = ['cola-lowrank-1-1', 'cola-linear-1-1']
        if task_name == 'ic':
            batch_size = ['256']
        elif model_names[0] == 'llama-2':
            batch_size = ['8']
        else:
            batch_size = ['32']
        control_name = [[data_names, model_names, [task_name], ft_name, batch_size]]
        controls = make_controls(control_name)
    elif mode == 'cola_dist_merge':
        data_names = ['dolly-15k']
        ft_name = ['cola-lowrank-1-1', 'cola-lowrank~linear-1-1']
        if model_names[0] == 'llama-2':
            batch_size = ['8']
        else:
            batch_size = ['32']
        dist_mode = ['col']
        control_name = [[data_names, model_names, [task_name], ft_name, batch_size, dist_mode]]
        controls = make_controls(control_name)
    else:
        raise ValueError('Not valid mode')
    return controls

ARGS = {
    'mode': 'full',
    'task_name': 'ic'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, dtype family float — a list 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
make_controls Ran
ran; our contract draft was wrong, not the code
diaoenmao/cola-collaborative-adaptation-with-gradient-learning/src/process.py
code served (permissive licence) · get_code("7e48cde2bcfbe423")
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 list, ndim 2, 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 itertools

num_experiments = 3

exp = [str(x) for x in list(range(num_experiments))]

def make_controls(control_name):
    control_names = []
    for i in range(len(control_name)):
        control_names.extend(list(('_'.join(x) for x in itertools.product(*control_name[i]))))
    controls = [exp] + [control_names]
    controls = list(itertools.product(*controls))
    return controls


# Setup: control_name is a list of lists of lists of strings
# The function iterates through control_name, and for each element,
# it takes the cartesian product of the inner lists and joins them with '_'

control_name = [
    [['a', 'b'], ['x', 'y']],
    [['p', 'q', 'r'], ['1', '2']]
]

ARGS = {
    'control_name': control_name
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, dtype family float — a list 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

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 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 2 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

A primary function of back-propagation is to compute both the gradient of hidden representations and parameters for optimization with gradient descent. Training large models requires high computational costs due to their vast parameter sizes. While Parameter-Efficient Fine-Tuning (PEFT) methods aim to train smaller auxiliary models to save computational space, they still present computational overheads, especially in Fine-Tuning as a Service (FTaaS) for numerous users. We introduce Collaborative Adaptation (ColA) with Gradient Learning (GL), a parameter-free, model-agnostic fine-tuning approach that decouples the computation of the gradient of hidden representations and parameters. In comparison to PEFT methods, ColA facilitates more cost-effective FTaaS by offloading the computation of the gradient to low-cost devices. We also provide a theoretical analysis of ColA and experimentally demonstrate that ColA can perform on par or better than existing PEFT methods on various benchmarks.

For agents

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

get_harvested_code_for_paper("2404.13844")
get_code_for_paper("2404.13844")
have("2404.13844")

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

curl https://syntology.ai/api/ran/2404.13844.json

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

[![Syntology run record](https://syntology.ai/api/ran/2404.13844.svg)](https://syntology.ai/paper/2404.13844)

Connect an agent — have() is free.