SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.04758 · 2020

Salvaging Federated Learning by Local Adaptation

arXiv · PDF · Open in the Atlas

Code that ran

3 of 8 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 3 have no contract verdict; 0 failures of our instrument; 5 no recorded run), measured 2026-11-21; 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/2002.04758.json · badge: /api/ran/2002.04758.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
ebagdasa/federated_adaptation canonical (classifier label)
classifier basis: census
3 ran · 4 no recorded run
SanaAwan5/FTLSimulation community (classifier label)
classifier basis not recorded on the link
0 ran · 1 no recorded run
FunctionStatusWhere it lives
create_table Ran
ran on a synthesized input
ebagdasa/federated_adaptation/utils/utils.py
code served (permissive licence) · get_code("e5d2ae1542e8323e")
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 str. 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)

# Analyzing the function:
# - create_table takes a params dict
# - It iterates over params.keys() and filters out 'folder_path'
# - It accesses params[x] for each key to get string representations
# - It joins these with ' | ' to create header, line, and values

# Create a simple params dict with various key-value pairs
params = {
    'model_name': 'ResNet50',
    'learning_rate': 0.001,
    'batch_size': 32,
    'epochs': 100,
    'accuracy': 0.95,
    'folder_path': '/path/to/folder'  # This will be filtered out
}

ARGS = {
    'params': params
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
fisher_matrix_diag Ran
ran on a synthesized input
ebagdasa/federated_adaptation/utils/utils.py
code served (permissive licence) · get_code("e655eb5653268285")
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)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x, hidden=None):
        return self.linear(x)
    
    def init_hidden(self, batch_size):
        return (torch.zeros(batch_size, 4), torch.zeros(batch_size, 4))

class Helper:
    def __init__(self):
        self.data_type = 'image'
        self.multi_gpu = False
        self.params = {
            'test_batch_size': 2,
            'bptt': 5
        }
        self.n_tokens = 3
    
    def get_batch(self, data_source, batch, evaluation=False):
        if isinstance(batch, tuple):
            data, targets = batch
            return data, targets
        else:
            data = torch.randn(2, 4)
            targets = torch.randint(0, 3, (2,))
            return data, targets

class SimpleDataset:
    def __init__(self):
        self.dataset = [(torch.randn(2, 4), torch.randint(0, 3, (2,))) for _ in range(3)]
    
    def __iter__(self):
        return iter(self.dataset)
    
    def __len__(self):
        return len(self.dataset)

target_model = SimpleModel()
helper = Helper()
data_source = SimpleDataset()
criterion = nn.CrossEntropyLoss()

ARGS = {
    'helper': helper,
    'data_source': data_source,
    'target_model': target_model,
    'criterion': criterion
}
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 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_word_list Ran
ran on a synthesized input
ebagdasa/federated_adaptation/utils/text_load.py
code served (permissive licence) · get_code("acda036b841b1af4")
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 list, ndim 1, 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 json
import re

filter_symbols = re.compile('[a-zA-Z]*')

def get_word_list(line, dictionary):
    splitted_words = json.loads(line.lower()).split()
    words = ['<bos>']
    for word in splitted_words:
        word = filter_symbols.search(word)[0]
        if len(word) > 1:
            if dictionary.word2idx.get(word, False):
                words.append(word)
            else:
                words.append('<unk>')
    words.append('<eos>')
    return words


class MockDictionary:
    def __init__(self):
        self.word2idx = {
            'hello': 1,
            'world': 2,
            'python': 3,
            'test': 4
        }


line = json.dumps("Hello world python test")
dictionary = MockDictionary()

ARGS = {
    'line': line,
    'dictionary': dictionary
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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 1, 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
criterion_ewc Not yet run
unverified (harvested, no recorded run)
ebagdasa/federated_adaptation/utils/utils.py
code served (permissive licence) · get_code("51c7799aa4830ba9")
How it ran

no invocation recorded.

eval_one_participant Not yet run
unverified (harvested, no recorded run)
ebagdasa/federated_adaptation/adapt.py
code served (permissive licence) · get_code("5ce7fe7ffc86a845")
How it ran

no invocation recorded.

test_globalmodel_local Not yet run
unverified (harvested, no recorded run)
ebagdasa/federated_adaptation/adapt.py
code served (permissive licence) · get_code("067dfdd931c5b6e2")
How it ran

no invocation recorded.

train Not yet run
unverified (harvested, no recorded run)
ebagdasa/federated_adaptation/training.py
code served (permissive licence) · get_code("2f2c95358bb2ba4b")
How it ran

no invocation recorded.

train Not yet run
unverified (harvested, no recorded run)
SanaAwan5/FTLSimulation/training.py
code served (permissive licence) · get_code("d1eabb8d72b78b3a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 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 8 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 2026-11-21.

Repositories linked to this paper

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

Abstract

Federated learning (FL) is a heavily promoted approach for training ML models on sensitive data, e.g., text typed by users on their smartphones. FL is expressly designed for training on data that are unbalanced and non-iid across the participants. To ensure privacy and integrity of the fedeated model, latest FL approaches use differential privacy or robust aggregation. We look at FL from the \emph{local} viewpoint of an individual participant and ask: (1) do participants have an incentive to participate in FL? (2) how can participants \emph{individually} improve the quality of their local models, without re-designing the FL framework and/or involving other participants? First, we show that on standard tasks such as next-word prediction, many participants gain no benefit from FL because the federated model is less accurate on their data than the models they can train locally on their own. Second, we show that differential privacy and robust aggregation make this problem worse by further destroying the accuracy of the federated model for many participants. Then, we evaluate three techniques for local adaptation of federated models: fine-tuning, multi-task learning, and knowledge distillation. We analyze where each is applicable and demonstrate that all participants benefit from local adaptation. Participants whose local models are poor obtain big accuracy improvements over conventional FL. Participants whose local models are better than the federated model\textemdash and who have no incentive to participate in FL today\textemdash improve less, but sufficiently to make the adapted federated model better than their local models.

For agents

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

get_harvested_code_for_paper("2002.04758")
get_code_for_paper("2002.04758")
have("2002.04758")

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

curl https://syntology.ai/api/ran/2002.04758.json

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

[![Syntology run record](https://syntology.ai/api/ran/2002.04758.svg)](https://syntology.ai/paper/2002.04758)

Connect an agent — have() is free.