SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2005.00115 · ACL · 2020

Learning to Faithfully Rationalize by Construction

Sarah Wiegreffe, Byron Wallace, Yuval Pinter, Sarthak Jain

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 8 have no contract verdict; 4 failures of our instrument; 2 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/2005.00115.json · badge: /api/ran/2005.00115.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
successar/FRESH canonical (classifier label)
classifier basis: census
12 ran · 2 no recorded run
FunctionStatusWhere it lives
annotations_from_jsonl Ran
ran; our contract draft was wrong, not the code
successar/FRESH/Datasets/utils.py
code served (permissive licence) · get_code("2843042d2c0e6f50")
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 tempfile
import os
from dataclasses import dataclass
from typing import Set, Tuple, Union, FrozenSet

@dataclass(eq=True, frozen=True)
class Evidence:
    text: Union[str, Tuple[int], Tuple[str]]
    docid: str
    start_token: int = -1
    end_token: int = -1
    start_sentence: int = -1
    end_sentence: int = -1

@dataclass(eq=True, frozen=True)
class Annotation:
    annotation_id: str
    query: Union[str, Tuple[int]]
    evidences: Union[Set[Tuple[Evidence]], FrozenSet[Tuple[Evidence]]]
    classification: str
    query_type: str = None
    docids: Set[str] = None

# Create a temporary JSONL file with valid data
temp_fd, temp_path = tempfile.mkstemp(suffix='.jsonl')
os.close(temp_fd)

# Build sample evidence groups
ev1 = Evidence(
    text="The quick brown fox",
    docid="doc1",
    start_token=0,
    end_token=4,
    start_sentence=0,
    end_sentence=1
)

ev2 = Evidence(
    text="jumps over the lazy dog",
    docid="doc2",
    start_token=0,
    end_token=5,
    start_sentence=0,
    end_sentence=1
)

# Create sample annotation records in JSONL format
annotations_data = [
    {
        "annotation_id": "ann_001",
        "query": "What does the fox do?",
        "evidences": [
            [
                {
                    "text": "The quick brown fox",
                    "docid": "doc1",
                    "start_token": 0,
                    "end_token": 4,
                    "start_sentence": 0,
                    "end_sentence": 1
                }
            ]
        ],
        "classification": "SUPPORTS",
        "query_type": "factual",
        "docids": ["doc1"]
    },
    {
        "annotation_id": "ann_002",
        "query": "What is the animal?",
        "evidences": [
            [
                {
                    "text": "jumps over the lazy dog",
                    "docid": "doc2",
                    "start_token": 0,
                    "end_token": 5,
                    "start_sentence": 0,
                    "end_sentence": 1
                }
            ]
        ],
        "classification": "REFUTES",
        "query_type": "factual",
        "docids": ["doc2"]
    }
]

# Write to temp file
with open(temp_path, 'w') as f:
    for record in annotations_data:
        f.write(json.dumps(record) + '\n')

ARGS = {
    'fp': temp_path
}
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
computer Ran
ran on a synthesized input
successar/FRESH/human_amt/compute_amt_statistics.py
code served (permissive licence) · get_code("70170b65c8a6c5e9")
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 float, 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

torch.manual_seed(0)

def computer(x):
    if x == 3 or x == 0:
        return 1.0
    elif x == 2 or x == 1:
        return 0.67
    else:
        raise Exception

ARGS = {
    "x": 3
}
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 float, 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
convert_to_tensor Ran
ran on a synthesized input
successar/FRESH/Rationale_Analysis/models/kuma.py
code served (permissive licence) · get_code("8f6eacf222a4c3f7")
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 1, dtype family float, shape (1,) 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

torch.manual_seed(0)

ARGS = {
    "x": 3.14
}
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 1, dtype family float, shape (1,) — shape (1,) 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
extract_raw_abstract Ran
ran on a synthesized input
successar/FRESH/Datasets/evinf/process_evidence_inference.py
code served (permissive licence) · get_code("8ddd3bcf97fbbcd0")
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

class Article:
    def get_abstract(self, param):
        return '<p>This is a test abstract</p><p>with multiple paragraphs</p>'

article = Article()

ARGS = {
    'article': article
}
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 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
kuma_mean Ran
ran on a synthesized input
successar/FRESH/Rationale_Analysis/models/kuma.py
code served (permissive licence) · get_code("aee97053c9473179")
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 1, dtype family float, shape (4,) 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

torch.manual_seed(0)

# kuma_mean calls kuma_moments with a, b, and n=1
# kuma_moments uses a and b in arithmetic operations and torch.lgamma
# So a and b should be torch tensors that support these operations

# Create small tensors for a and b parameters
a = torch.tensor([0.5, 1.5, 2.0, 3.0], dtype=torch.float32)
b = torch.tensor([0.8, 1.2, 2.5, 1.8], dtype=torch.float32)

ARGS = {
    "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 1, dtype family float, shape (4,) — shape (4,) 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
kuma_moments Ran
ran; our contract draft was wrong, not the code
successar/FRESH/Rationale_Analysis/models/kuma.py
pointer only (licence: MIT); the stored copy get_code serves, from bhargaviparanjape/explainable_qa, is pointer only (licence: NONE) · get_code("b6c1438a7ae7e272")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from bhargaviparanjape/explainable_qa, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

load_datasets Ran
ran; our contract draft was wrong, not the code
successar/FRESH/Datasets/utils.py
code served (permissive licence) · get_code("e1878c4a3bf022ae")
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. 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 json
import os
import tempfile
from dataclasses import dataclass
from typing import Set, Tuple, Union, FrozenSet

@dataclass(eq=True, frozen=True)
class Evidence:
    text: Union[str, Tuple[int], Tuple[str]]
    docid: str
    start_token: int = -1
    end_token: int = -1
    start_sentence: int = -1
    end_sentence: int = -1

@dataclass(eq=True, frozen=True)
class Annotation:
    annotation_id: str
    query: Union[str, Tuple[int]]
    evidences: Union[Set[Tuple[Evidence]], FrozenSet[Tuple[Evidence]]]
    classification: str
    query_type: str = None
    docids: Set[str] = None

# Create a temporary directory with train.jsonl, val.jsonl, test.jsonl
temp_dir = tempfile.mkdtemp()

# Create sample Evidence objects
evidence1 = Evidence(
    text="This is evidence text",
    docid="doc1",
    start_token=0,
    end_token=5,
    start_sentence=0,
    end_sentence=1
)

evidence2 = Evidence(
    text="Another evidence",
    docid="doc2",
    start_token=1,
    end_token=3,
    start_sentence=0,
    end_sentence=1
)

# Create sample Annotation objects
annotation1 = {
    "annotation_id": "ann1",
    "query": "What is this?",
    "evidences": [[
        {
            "text": "Evidence text 1",
            "docid": "doc1",
            "start_token": 0,
            "end_token": 5,
            "start_sentence": 0,
            "end_sentence": 1
        }
    ]],
    "classification": "SUPPORTS",
    "query_type": "factoid",
    "docids": ["doc1"]
}

annotation2 = {
    "annotation_id": "ann2",
    "query": "Why?",
    "evidences": [[
        {
            "text": "Evidence text 2",
            "docid": "doc2",
            "start_token": 1,
            "end_token": 3,
            "start_sentence": 0,
            "end_sentence": 1
        }
    ]],
    "classification": "REFUTES",
    "query_type": "factoid",
    "docids": ["doc2"]
}

# Write train.jsonl
with open(os.path.join(temp_dir, 'train.jsonl'), 'w') as f:
    f.write(json.dumps(annotation1) + '\n')
    f.write(json.dumps(annotation2) + '\n')

# Write val.jsonl
with open(os.path.join(temp_dir, 'val.jsonl'), 'w') as f:
    f.write(json.dumps(annotation1) + '\n')

# Write test.jsonl
with open(os.path.join(temp_dir, 'test.jsonl'), 'w') as f:
    f.write(json.dumps(annotation2) + '\n')

ARGS = {
    'data_dir': temp_dir
}
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 tuple — a tuple output has no array shape of its own; none is 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
load_jsonl Ran
ran; our contract draft was wrong, not the code
successar/FRESH/Datasets/utils.py
code served (permissive licence) · get_code("49077456b34beefd")
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 tempfile
import os

# Create a temporary JSONL file with test data
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.jsonl")

# Write test data to the file
test_data = [
    {"id": 1, "text": "hello world"},
    {"id": 2, "text": "foo bar"},
    {"id": 3, "name": "test", "value": 42}
]

with open(temp_file, 'w') as f:
    for item in test_data:
        json.dump(item, f)
        f.write('\n')

# Define ARGS
ARGS = {
    "fp": temp_file
}
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
mapper Ran
ran on a synthesized input
successar/FRESH/human_amt/compute_amt_statistics.py
code served (permissive licence) · get_code("9a0ae9e831dc0d57")
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 float, 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

def mapper(x):
    if x == 'Very Difficult':
        return 1.0
    elif x == 'Difficult':
        return 2.0
    elif x == 'Neutral':
        return 3.0
    elif x == 'Easy':
        return 4.0
    elif x == 'Very Easy':
        return 5.0

# mapper takes a single string parameter x and returns a float
# Test with one of the valid difficulty strings
ARGS = {
    'x': 'Very Difficult'
}
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 float, 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
samplemax Ran
ran on a synthesized input
successar/FRESH/Rationale_Analysis/experiments/allentune_plot.py
code served (permissive licence) · get_code("091c323ea322b6ab")
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 numpy as np
import scipy.special

ARGS = {
    'validation_performance': [0.72, 0.65, 0.81, 0.58, 0.91],
    'with_replacement': True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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
td_format Ran
ran on a synthesized input
successar/FRESH/Rationale_Analysis/experiments/allentune_plot.py
code served (permissive licence) · get_code("44f8471c142ddd5c")
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.

from datetime import timedelta

# The function expects a timedelta object with a total_seconds() method
td_object = timedelta(days=2, hours=3, minutes=45, seconds=30)

ARGS = {
    'td_object': td_object
}
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 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
vote Ran
ran on a synthesized input
successar/FRESH/human_amt/compute_amt_statistics.py
code served (permissive licence) · get_code("cd1019632838faed")
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

torch.manual_seed(0)

# The function takes a single parameter 'x' and checks if it equals 2, 3, 0, or 1
# It returns 1 if x is 2 or 3, returns 0 if x is 0 or 1, otherwise raises an Exception
# So x should be an integer in the range [0, 1, 2, 3]

ARGS = {
    "x": 2
}
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 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
generate_report Not yet run
unverified (harvested, no recorded run)
successar/FRESH/Rationale_Analysis/experiments/allentune_plot.py
code served (permissive licence) · get_code("66de11959336aeb2")
How it ran

no invocation recorded.

main Not yet run
unverified (harvested, no recorded run)
successar/FRESH/Datasets/get_dataset_statistics.py
code served (permissive licence) · get_code("ff594b5da33b1042")
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, 6 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 1 the same month, 3 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.

FunctionAlso present inEarliest-datedOthers named
annotations_from_jsonl 3 other papers
1 dated earlier, 2 later
1911.03429 (2019-11)
ERASER: A Benchmark to Evaluate Rationalized NLP Models
2310.16316 (2023-10)
Sum-of-Parts: Self-Attributing Neural Networks with End-to-End Learning of Feature Groups
2024.findings-acl.307 (2024 (year only))
Plausible Extractive Rationalization through Semi-Supervised Entailment Signal
load_datasets 3 other papers
1 dated earlier, 2 later
1911.03429 (2019-11)
ERASER: A Benchmark to Evaluate Rationalized NLP Models
2310.16316 (2023-10)
Sum-of-Parts: Self-Attributing Neural Networks with End-to-End Learning of Feature Groups
2024.findings-acl.307 (2024 (year only))
Plausible Extractive Rationalization through Semi-Supervised Entailment Signal
load_jsonl 3 other papers
1 dated earlier, 2 later
1911.03429 (2019-11)
ERASER: A Benchmark to Evaluate Rationalized NLP Models
2310.16316 (2023-10)
Sum-of-Parts: Self-Attributing Neural Networks with End-to-End Learning of Feature Groups
2024.findings-acl.307 (2024 (year only))
Plausible Extractive Rationalization through Semi-Supervised Entailment Signal
kuma_moments 1 other paper
1 the same month
2005.00652 (2020-05)
An Information Bottleneck Approach for Controlling Conciseness in Rationale Extraction
samplemax 1 other paper
1 dated earlier
1909.03004 (2019-09)
Show Your Work: Improved Reporting of Experimental Results
td_format 1 other paper
1 dated earlier
1909.03004 (2019-09)
Show Your Work: Improved Reporting of Experimental Results

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

In many settings it is important for one to be able to understand why a model made a particular prediction. In NLP this often entails extracting snippets of an input text 'responsible for' corresponding model output; when such a snippet comprises tokens that indeed informed the model's prediction, it is a faithful explanation. In some settings, faithfulness may be critical to ensure transparency. Lei et al. (2016) proposed a model to produce faithful rationales for neural text classification by defining independent snippet extraction and prediction modules. However, the discrete selection over input tokens performed by this method complicates training, leading to high variance and requiring careful hyperparameter tuning. We propose a simpler variant of this approach that provides faithful explanations by construction. In our scheme, named FRESH, arbitrary feature importance scores (e.g., gradients from a trained model) are used to induce binary labels over token inputs, which an extractor can be trained to predict. An independent classifier module is then trained exclusively on snippets provided by the extractor; these snippets thus constitute faithful explanations, even if the classifier is arbitrarily complex. In both automatic and manual evaluations we find that variants of this simple framework yield predictive performance superior to 'end-to-end' approaches, while being more general and easier to train. 1

For agents

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

get_harvested_code_for_paper("2005.00115")
get_code_for_paper("2005.00115")
have("2005.00115")

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

curl https://syntology.ai/api/ran/2005.00115.json

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

[![Syntology run record](https://syntology.ai/api/ran/2005.00115.svg)](https://syntology.ai/paper/2005.00115)

Connect an agent — have() is free.