SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2401.00396 · 2024

RAGTruth: A Hallucination Corpus for Developing Trustworthy Retrieval-Augmented Language Models

arXiv · PDF · Open in the Atlas

Code that ran

3 of 5 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; 3 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/2401.00396.json · badge: /api/ran/2401.00396.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
particlemedia/ragtruth canonical (classifier label)
classifier basis: census
3 ran · 2 no recorded run
FunctionStatusWhere it lives
get_id Ran
ran; our contract draft was wrong, not the code
particlemedia/ragtruth/baseline/prepare_dataset.py
code served (permissive licence) · get_code("550abe4b00a769a6")
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

torch.manual_seed(0)

item = {
    'id': 'item_001',
    'sentence_id': 'sent_42',
    'model': 'bert_base'
}

ARGS = {
    'item': item
}
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
get_json_data Ran
ran; our contract draft was wrong, not the code
particlemedia/ragtruth/baseline/prepare_dataset.py
code served (permissive licence) · get_code("80c5ca0bb7603aac")
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 pandas as pd
import torch

torch.manual_seed(0)

# Create a minimal DataFrame that matches the function's expectations
data = pd.DataFrame({
    'prompt': ['What is X?', 'Summarize Y'],
    'response': ['Response A', 'Response B'],
    'labels': [
        [
            {'label_type': 'baseless info', 'text': 'false claim 1', 'start': 0},
            {'label_type': 'conflict', 'text': 'conflicting info', 'start': 5}
        ],
        [
            {'label_type': 'baseless', 'text': 'unsupported', 'start': 2}
        ]
    ],
    'task_type': ['QA', 'Summary'],
    'source_info': [
        {'passages': ['passage1', 'passage2'], 'question': 'What is X?'},
        'Summary source text'
    ]
})

ARGS = {
    'data': data
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas, torch — 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
read_ragtruth_split Ran
ran; our contract draft was wrong, not the code
particlemedia/ragtruth/baseline/prepare_dataset.py
code served (permissive licence) · get_code("ea94977c202a27c9")
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 DataFrame. 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 pandas as pd
import tempfile
import os

# Create a temporary directory for test data
tmpdir = tempfile.mkdtemp()

# Create response.jsonl with test data
response_data = [
    {'source_id': 1, 'split': 'train', 'quality': 'good', 'text': 'response1'},
    {'source_id': 2, 'split': 'test', 'quality': 'good', 'text': 'response2'},
    {'source_id': 3, 'split': 'test', 'quality': 'bad', 'text': 'response3'},
    {'source_id': 4, 'split': 'test', 'quality': 'good', 'text': 'response4'},
]
response_df = pd.DataFrame(response_data)
response_df.to_json(os.path.join(tmpdir, 'response.jsonl'), orient='records', lines=True)

# Create source_info.jsonl with test data
source_info_data = [
    {'source_id': 1, 'source': 'wiki', 'url': 'http://example.com/1'},
    {'source_id': 2, 'source': 'wiki', 'url': 'http://example.com/2'},
    {'source_id': 3, 'source': 'news', 'url': 'http://example.com/3'},
    {'source_id': 4, 'source': 'wiki', 'url': 'http://example.com/4'},
]
source_info_df = pd.DataFrame(source_info_data)
source_info_df.to_json(os.path.join(tmpdir, 'source_info.jsonl'), orient='records', lines=True)

ARGS = {
    'ragtruth_dir': tmpdir,
    'split': 'test',
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DataFrame — a DataFrame output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DataFrame output the recorded ndim 2 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
process_dialog Not yet run
unverified (harvested, no recorded run)
particlemedia/ragtruth/baseline/dataset.py
code served (permissive licence) · get_code("d103105e08e801d7")
How it ran

no invocation recorded.

process_dialog_to_single_turn Not yet run
unverified (harvested, no recorded run)
particlemedia/ragtruth/baseline/dataset.py
code served (permissive licence) · get_code("9da9fe51afe5a3bf")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 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 5 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

Retrieval-augmented generation (RAG) has become a main technique for alleviating hallucinations in large language models (LLMs). Despite the integration of RAG, LLMs may still present unsupported or contradictory claims to the retrieved contents. In order to develop effective hallucination prevention strategies under RAG, it is important to create benchmark datasets that can measure the extent of hallucination. This paper presents RAGTruth, a corpus tailored for analyzing word-level hallucinations in various domains and tasks within the standard RAG frameworks for LLM applications. RAGTruth comprises nearly 18,000 naturally generated responses from diverse LLMs using RAG. These responses have undergone meticulous manual annotations at both the individual cases and word levels, incorporating evaluations of hallucination intensity. We not only benchmark hallucination frequencies across different LLMs, but also critically assess the effectiveness of several existing hallucination detection methodologies. Furthermore, we show that using a high-quality dataset such as RAGTruth, it is possible to finetune a relatively small LLM and achieve a competitive level of performance in hallucination detection when compared to the existing prompt-based approaches using state-of-the-art large language models such as GPT-4.

For agents

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

get_harvested_code_for_paper("2401.00396")
get_code_for_paper("2401.00396")
have("2401.00396")

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

curl https://syntology.ai/api/ran/2401.00396.json

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

[![Syntology run record](https://syntology.ai/api/ran/2401.00396.svg)](https://syntology.ai/paper/2401.00396)

Connect an agent — have() is free.