SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2104.05344 · ICLR · 2021

How Sensitive are Meta-Learners to Dataset Imbalance?

Amos Storkey, Massimiliano Patacchiola, Sen Wang, Mateusz Ochal, Jose Vazquez

arXiv · PDF · Open in the Atlas

Code that ran

2 of 3 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; 1 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/2104.05344.json · badge: /api/ran/2104.05344.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
mattochal/imbalanced_fsl_public canonical (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
FunctionStatusWhere it lives
flatten_dict Ran
ran; our contract draft was wrong, not the code
mattochal/imbalanced_fsl_public/generator.py
code served (permissive licence) · get_code("0b2f8809055e38f5")
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 numpy as np

# Analyze the function: flatten_dict takes a dictionary and flattens it recursively
# It uses:
# - d: a dict to flatten (can contain nested dicts)
# - prefix: optional string prefix for keys
# - seperator: string separator (default '.')
# - value_map: a callable that maps values (though it's defined but never used in the function body)

# Create a nested dictionary structure to test the flattening
nested_dict = {
    'a': 1,
    'b': {
        'c': 2,
        'd': {
            'e': 3
        }
    },
    'f': 'test_string'
}

# Define ARGS with all parameters
ARGS = {
    'd': nested_dict,
    'prefix': None,
    'seperator': '.',
    'value_map': lambda x: x
}
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 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
unpack Ran
ran; our contract draft was wrong, not the code
mattochal/imbalanced_fsl_public/generator.py
code served (permissive licence) · get_code("1bdfd97bb9483c1b")
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

torch.manual_seed(0)

# The function 'unpack' takes a dictionary 'comb' and unpacks it
# It looks for keys and values that are both tuples
# If found, it zips them together to create a new dictionary
# Otherwise, it keeps the key-value pair as is

# Construct a valid argument for the function
comb = {
    # Regular key-value pair (not a tuple key)
    "regular_key": "regular_value",
    # Tuple key with tuple value - these will be unpacked
    ("key1", "key2", "key3"): ("value1", "value2", "value3"),
    # Another regular pair
    "another_key": 42,
    # Another tuple pair with different lengths
    ("a", "b"): (10, 20),
}

ARGS = {
    "comb": comb
}
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
substitute_hyperparams Not yet run
unverified (harvested, no recorded run)
mattochal/imbalanced_fsl_public/generator.py
code served (permissive licence) · get_code("89ea0294d48155c8")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 0 later) and 0 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
flatten_dict 1 other paper
1 dated earlier
2101.02523 (2021-01)
Few-Shot Learning with Class Imbalance
substitute_hyperparams 1 other paper
1 dated earlier
2101.02523 (2021-01)
Few-Shot Learning with Class Imbalance
unpack 1 other paper
1 dated earlier
2101.02523 (2021-01)
Few-Shot Learning with Class Imbalance

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

Meta-Learning (ML) has proven to be a useful tool for training Few-Shot Learning (FSL) algorithms by exposure to batches of tasks sampled from a meta-dataset. However, the standard training procedure overlooks the dynamic nature of the real-world where object classes are likely to occur at different frequencies. While it is generally understood that imbalanced tasks harm the performance of supervised methods, there is no significant research examining the impact of imbalanced meta-datasets on the FSL evaluation task. This study exposes the magnitude and extent of this problem. Our results show that ML methods are more robust against meta-dataset imbalance than imbalance at the task-level with a similar imbalance ratio (ρ < 20), with the effect holding even in long-tail datasets under a larger imbalance (ρ = 65). Overall, these results highlight an implicit strength of ML algorithms, capable of learning generalizable features under dataset imbalance and domain-shift. The code to reproduce the experiments is released under an open-source license 1 .

For agents

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

get_harvested_code_for_paper("2104.05344")
get_code_for_paper("2104.05344")
have("2104.05344")

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

curl https://syntology.ai/api/ran/2104.05344.json

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

[![Syntology run record](https://syntology.ai/api/ran/2104.05344.svg)](https://syntology.ai/paper/2104.05344)

Connect an agent — have() is free.