SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2501.19010 · NAACL · 2025

DyPCL: Dynamic Phoneme-level Contrastive Learning for Dysarthric Speech Recognition

Jungseul Ok, Wonjun Lee, Heejin Do, Yunsu Kim, Gary Lee, Solee Im

arXiv · PDF · Open in the Atlas

Code that ran

12 of 14 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 12 have no contract verdict; 0 failures of our instrument; 2 no recorded run), measured 2026-11-22; 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/2501.19010.json · badge: /api/ran/2501.19010.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
dmort27/panphon canonical (classifier label)
classifier basis not recorded on the link
12 ran · 2 no recorded run
FunctionStatusWhere it lives
dogol_leven_dist Ran
ran on a synthesized input
dmort27/panphon/panphon/bin/align_wordlists.py
code served (permissive licence) · get_code("ce3334ed30edebac")
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 int, ndim 0, dtype family int, 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
import numpy as np
import Levenshtein

torch.manual_seed(0)

class MockDistanceMapper:
    """Mock distance object with map_to_dogol_prime method"""
    def map_to_dogol_prime(self, x):
        """Convert input to a string representation for Levenshtein distance"""
        if isinstance(x, str):
            return x
        elif isinstance(x, (list, tuple)):
            return "".join(str(item) for item in x)
        elif isinstance(x, torch.Tensor):
            return "".join(str(int(v)) for v in x.flatten().tolist())
        else:
            return str(x)

dist = MockDistanceMapper()
a = "hello"
b = "hallo"

ARGS = {
    "dist": dist,
    "a": a,
    "b": b,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
Levenshtein, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, 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
feature_hamming_dist Ran
ran on a synthesized input
dmort27/panphon/panphon/bin/align_wordlists.py
code served (permissive licence) · get_code("0c005def5a2a6c24")
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
import torch.nn as nn

torch.manual_seed(0)

class MockDistanceModule(nn.Module):
    def __init__(self):
        super().__init__()
    
    def feature_edit_distance(self, a, b):
        # Return a dummy edit distance value
        # The function just calls this method and returns its result
        return torch.tensor(5.0)

dist = MockDistanceModule()
a = torch.randn(2, 4)
b = torch.randn(2, 4)

ARGS = {
    "dist": dist,
    "a": a,
    "b": b
}
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
flip Ran
ran on a synthesized input
dmort27/panphon/panphon/permissive.py
code served (permissive licence) · get_code("de4fc622bf3691c5")
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 2, 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
import numpy as np

torch.manual_seed(0)

# The function flip takes a parameter 's' which should be iterable
# and returns a list of tuples where each (a, b) pair is reversed to (b, a)
# This is a simple sequence of tuples that needs to be flipped

s = [(1, 2), (3, 4), (5, 6)]

ARGS = {
    "s": s
}
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 list, ndim 2, dtype family int — 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
fts Ran
ran on a synthesized input
dmort27/panphon/panphon/_panphon.py
code served (permissive licence) · get_code("1f5a3c75c78cb558")
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 set. 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 regex as re
from typing import Tuple, Set

# The function takes a single parameter 's' which is a string
# containing feature specifications with +/- prefixes followed by alphabetical sequences
# Example: "+son -syl 0cor" should parse to {('+', 'son'), ('-', 'syl'), ('0', 'cor')}

ARGS = {
    's': "+son -syl 0cor +voi -nas"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
regex — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type set — a set output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a set 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
ftstr2dict Ran
ran on a synthesized input
dmort27/panphon/panphon/distance.py
code served (permissive licence) · get_code("277f97e8c5873b8c")
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 regex as re

def ftstr2dict(ftstr: str):
    """
    Parse a feature string into a dictionary.
    The function uses:
    - re.finditer() to find patterns matching '([-0+])(\\w+)'
    - m.groups() to extract matched groups
    - Dictionary lookup with keys '-', '0', '+'
    """
    fts = {}
    for m in re.finditer('([-0+])(\\w+)', ftstr):
        v, k = m.groups()
        fts[k] = {'-': -1, '0': 0, '+': 1}[v]
    return fts


# Create test inputs
# The function expects a string with feature patterns like "-feature1+feature2"
ftstr_input = "-alpha+beta0gamma"

ARGS = {
    "ftstr": ftstr_input
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
regex — 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
levenshtein_dist Ran
ran on a synthesized input
dmort27/panphon/panphon/bin/align_wordlists.py
code served (permissive licence) · get_code("409384dc1505919f")
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 int, ndim 0, dtype family int, 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 Levenshtein

torch_available = False
try:
    import torch
    torch_available = True
    torch.manual_seed(0)
except ImportError:
    pass

# The function levenshtein_dist takes three parameters: _, a, b
# The first parameter (_) is unused
# Parameters a and b are strings passed to Levenshtein.distance()
# which computes the Levenshtein distance between two strings

ARGS = {
    "_": None,
    "a": "hello",
    "b": "hallo"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
Levenshtein, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, 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
pat Ran
ran on a synthesized input
dmort27/panphon/panphon/_panphon.py
code served (permissive licence) · get_code("2a594bf8d80114e7")
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 regex as re

# Test case 1: Simple feature pattern with one segment
p1 = "[+voice -nasal 0rounded]"

# Test case 2: Multiple segments
p2 = "[+high -back][+voiced]"

# Test case 3: Empty or minimal patterns
p3 = "[]"

# Test case 4: Complex pattern with multiple features
p4 = "[+consonant -vowel 0nasal][+high +front -rounded][+voiced]"

ARGS = {
    "p": p4
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
regex — 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
update_ft_set Ran
ran on a synthesized input
dmort27/panphon/panphon/permissive.py
code served (permissive licence) · get_code("5ba9dee80f06cc42")
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 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 torch
import numpy as np

def flip(s):
    return [(b, a) for a, b in s]

def update_ft_set(seg, dia):
    seg = dict(flip(seg))
    seg.update(dia)
    return flip(set(seg.items()))

# Analyzing the function:
# - seg is passed to flip(), which expects an iterable of 2-tuples
# - The result is converted to dict, so seg should be a sequence of (key, value) pairs
# - dia is used with seg.update(), so dia should be dict-like
# - flip() is called on the result of set(seg.items()), which produces tuples

# seg: a sequence of tuples (key, value)
seg_value = [('a', 1), ('b', 2), ('c', 3)]

# dia: a dictionary to update seg with
dia_value = {'d': 4, 'e': 5}

ARGS = {
    'seg': seg_value,
    'dia': dia_value
}
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 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
vector_to_tuple Ran
ran on a synthesized input
dmort27/panphon/panphon/panphonvec.py
code served (permissive licence) · get_code("40ff49612ac3ccee")
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 numpy as np

# Create a small numpy array with integer values
vector = np.array([1, 2, 3, 4], dtype=int)

ARGS = {
    "vector": vector
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — 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
word2array Ran
ran on a synthesized input
dmort27/panphon/panphon/_panphon.py
code served (permissive licence) · get_code("f8e148176e389e52")
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 ndarray, ndim 2, dtype family int, shape (3, 3) 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 numpy as np
from typing import List, Tuple

# Construct the arguments for word2array

# ft_names: List of feature names as strings
ft_names = ['feature_1', 'feature_2', 'feature_3']

# word: List of lists of (value, feature) tuples
# Each inner list represents a segment with feature tuples (value, feature_name)
# The function expects tuples where first element is the value (+, -, or 0)
# and second element is the feature name
word = [
    [
        ('+', 'feature_1'),
        ('-', 'feature_2'),
        ('0', 'feature_3'),
    ],
    [
        ('0', 'feature_1'),
        ('+', 'feature_2'),
        ('-', 'feature_3'),
    ],
    [
        ('-', 'feature_1'),
        ('0', 'feature_2'),
        ('+', 'feature_3'),
    ],
]

ARGS = {
    'ft_names': ft_names,
    'word': word,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family int, shape (3, 3) — shape (3, 3) 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
xsampaopt Ran
ran on a synthesized input
dmort27/panphon/panphon/distance.py
code served (permissive licence) · get_code("78b14be49ee5b8e9")
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 function. 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
from typing import Callable, Any, Dict

def xsampaopt(f: Callable) -> Callable:
    def wrapper(*args: Any, **kwargs: Any) -> Any:
        if 'xsampa' in kwargs and kwargs['xsampa']:
            self, source, target = args
            source = self.xs.convert(source)
            target = self.xs.convert(target)
            args = (self, source, target)
        return f(*args, **kwargs)
    return wrapper


class XSampaConverter:
    def convert(self, text):
        return text.upper()


class DummySelf:
    def __init__(self):
        self.xs = XSampaConverter()


def dummy_function(self, source, target):
    return f"source: {source}, target: {target}"


decorated_function = xsampaopt(dummy_function)

ARGS = {
    'f': decorated_function
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
zerodiviszero Ran
ran on a synthesized input
dmort27/panphon/panphon/distance.py
code served (permissive licence) · get_code("66168504601bef77")
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 function. 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 function that raises ZeroDivisionError
def division_function(a, b):
    return a / b

# Create a function that works without error
def normal_function(x, y):
    return x + y

# Test with a function that raises ZeroDivisionError
def zero_div_raiser():
    return 10 / 0

# Test with a function that works fine
def regular_func():
    return 42

ARGS = {
    "f": zero_div_raiser
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
generate_feature_vectors Not yet run
unverified (harvested, no recorded run)
dmort27/panphon/panphon/panphonvec.py
code served (permissive licence) · get_code("286770a9471a65cf")
How it ran

no invocation recorded.

generate_modifiers Not yet run
unverified (harvested, no recorded run)
dmort27/panphon/panphon/panphonvec.py
code served (permissive licence) · get_code("0bacc4d8991f7d99")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 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 14 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

Dysarthric speech recognition often suffers from performance degradation due to the intrinsic diversity of dysarthric severity and extrinsic disparity from normal speech. To bridge these gaps, we propose a Dynamic Phonemelevel Contrastive Learning (DyPCL) method, which leads to obtaining invariant representations across diverse speakers. We decompose the speech utterance into phoneme segments for phoneme-level contrastive learning, leveraging dynamic connectionist temporal classification alignment. Unlike prior studies focusing on utterance-level embeddings, our granular learning allows discrimination of subtle parts of speech. In addition, we introduce dynamic curriculum learning, which progressively transitions from easy negative samples to difficultto-distinguish negative samples based on phonetic similarity of the phoneme. Our approach to training by difficulty levels alleviates the inherent variability of speakers, better identifying challenging speeches. Evaluated on the UASpeech dataset, DyPCL outperforms baseline models, achieving an average 22.10% relative reduction in word error rate (WER) across the overall dysarthria group.

For agents

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

get_harvested_code_for_paper("2501.19010")
get_code_for_paper("2501.19010")
have("2501.19010")

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

curl https://syntology.ai/api/ran/2501.19010.json

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

[![Syntology run record](https://syntology.ai/api/ran/2501.19010.svg)](https://syntology.ai/paper/2501.19010)

Connect an agent — have() is free.