SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2209.11222 · NeurIPS · 2022

Concept Activation Regions: A Generalized Framework For Concept-Based Explanations

Mihaela Van Der Schaar, Jonathan Crabbé

arXiv · PDF · Open in the Atlas

Code that ran

2 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 failures of our instrument; 5 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/2209.11222.json · badge: /api/ran/2209.11222.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
vanderschaarlab/CARs canonical (classifier label)
classifier basis not recorded on the link
1 ran · 4 no recorded run
jonathancrabbe/cars canonical (classifier label)
classifier basis: classifier_2025
1 ran · 1 no recorded run
FunctionStatusWhere it lives
CAR Ran
ran on a synthesized input · constructed an object; no result was computed
jonathancrabbe/cars/explanations/concept.py
code served (permissive licence) · get_code("5147ca73eac1472f")
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 an instance of its own class, CAR: the run constructed the object and did not call it, so no result was computed. 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
from sklearn.svm import SVC

torch.manual_seed(0)
np.random.seed(0)

# Create small test data
concept_reps = np.random.randn(10, 4).astype(np.float32)
concept_labels = np.array([0, 1, 0, 1, 0, 1, 0, 1, 0, 1], dtype=np.int32)

device = torch.device('cpu')

ARGS = {
    'device': device,
    'batch_size': 50,
    'kernel': 'rbf',
    'kernel_width': None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, optuna, sklearn, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type CAR — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a CAR 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
handle_integer_input Ran
ran on a synthesized input
vanderschaarlab/CARs/src/cars/models/mnist.py
code served (permissive licence) · get_code("3c0bdb3ee24fc417")
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 1, dtype family int; 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

torch.manual_seed(0)

# Test case 1: integer input
input_int = 5
desired_len_int = 3

# Test case 2: tuple input with correct length
input_tuple = (2, 3, 4)
desired_len_tuple = 3

ARGS = {
    "input": input_int,
    "desired_len": desired_len_int
}
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 tuple, ndim 1, dtype family int — 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
ConceptExplainer Not yet run
unverified (harvested, no recorded run)
jonathancrabbe/cars/explanations/concept.py
code served (permissive licence) · get_code("dcc35eab5bab9ae3")
How it ran

no invocation recorded.

generate_mnist_concept_dataset Not yet run
unverified (harvested, no recorded run)
vanderschaarlab/CARs/src/cars/utils/dataset.py
code served (permissive licence) · get_code("84c29fef946ee046")
How it ran

no invocation recorded.

get_dataloader Not yet run
unverified (harvested, no recorded run)
vanderschaarlab/CARs/src/cars/models/mnist.py
code served (permissive licence) · get_code("469a0e6c079bb107")
How it ran

no invocation recorded.

load_cub_data Not yet run
unverified (harvested, no recorded run)
vanderschaarlab/CARs/src/cars/utils/dataset.py
code served (permissive licence) · get_code("94c74e147491e689")
How it ran

no invocation recorded.

load_mnist Not yet run
unverified (harvested, no recorded run)
vanderschaarlab/CARs/src/cars/models/mnist.py
code served (permissive licence) · get_code("46ff66ae7e7ac875")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 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 7 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

Concept-based explanations permit to understand the predictions of a deep neural network (DNN) through the lens of concepts specified by users. Existing methods assume that the examples illustrating a concept are mapped in a fixed direction of the DNN's latent space. When this holds true, the concept can be represented by a concept activation vector (CAV) pointing in that direction. In this work, we propose to relax this assumption by allowing concept examples to be scattered across different clusters in the DNN's latent space. Each concept is then represented by a region of the DNN's latent space that includes these clusters and that we call concept activation region (CAR). To formalize this idea, we introduce an extension of the CAV formalism that is based on the kernel trick and support vector classifiers. This CAR formalism yields global concept-based explanations and local conceptbased feature importance. We prove that CAR explanations built with radial kernels are invariant under latent space isometries. In this way, CAR assigns the same explanations to latent spaces that have the same geometry. We further demonstrate empirically that CARs offer (1) more accurate descriptions of how concepts are scattered in the DNN's latent space; (2) global explanations that are closer to human concept annotations and (3) concept-based feature importance that meaningfully relate concepts with each other. Finally, we use CARs to show that DNNs can autonomously rediscover known scientific concepts, such as the prostate cancer grading system. 36th Conference on Neural Information Processing Systems (NeurIPS 2022).

For agents

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

get_harvested_code_for_paper("2209.11222")
get_code_for_paper("2209.11222")
have("2209.11222")

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

curl https://syntology.ai/api/ran/2209.11222.json

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

[![Syntology run record](https://syntology.ai/api/ran/2209.11222.svg)](https://syntology.ai/paper/2209.11222)

Connect an agent — have() is free.