SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2110.08466 · 2021

On the Safety of Conversational Models: Taxonomy, Dataset, and Benchmark

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 2 have no contract verdict; 0 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/2110.08466.json · badge: /api/ran/2110.08466.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
thu-coai/diasafety canonical (classifier label)
classifier basis not recorded on the link
2 ran · 1 no recorded run
FunctionStatusWhere it lives
get_loader Ran
ran on a synthesized input
thu-coai/diasafety/codes/mix_train.py
code served (permissive licence) · get_code("3349be060ae9fe89")
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 torch
from transformers import AutoTokenizer

torch.manual_seed(0)

# Create a minimal tokenizer stand-in that mimics the behavior
class MockTokenizer:
    def __call__(self, text, text_pair, return_tensors, truncation, padding, max_length):
        batch_size = len(text)
        return {
            'input_ids': torch.randint(0, 100, (batch_size, max_length)),
            'attention_mask': torch.ones(batch_size, max_length, dtype=torch.long),
            'token_type_ids': torch.zeros(batch_size, max_length, dtype=torch.long),
        }

# Create a small dataset
dataset = [
    {
        'context': 'This is context 1',
        'response': 'This is response 1',
        'category': 'Toxicity Agreement',
        'label': 'Unsafe'
    },
    {
        'context': 'This is context 2',
        'response': 'This is response 2',
        'category': 'Unauthorized Expertise',
        'label': 'Safe'
    },
    {
        'context': 'This is context 3',
        'response': 'This is response 3',
        'category': 'Offending User',
        'label': 'Unsafe'
    },
    {
        'context': 'This is context 4',
        'response': 'This is response 4',
        'category': 'Biased Opinion',
        'label': 'Safe'
    },
]

tokenizer = MockTokenizer()

ARGS = {
    'dataset': dataset,
    'tokenizer': tokenizer,
    'batchsize': 2,
    'padsize': 128,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm, transformers — 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
get_loader_resp Ran
ran on a synthesized input
thu-coai/diasafety/codes/mix_train.py
code served (permissive licence) · get_code("0862498ca3211ec4")
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 torch
from transformers import AutoTokenizer

torch.manual_seed(0)

# Create a minimal tokenizer stand-in that implements what the function uses
class MinimalTokenizer:
    def __call__(self, text, return_tensors='pt', truncation=True, padding='max_length', max_length=256):
        # Simulate tokenizer output with input_ids and attention_mask
        batch_size = len(text) if isinstance(text, list) else 1
        input_ids = torch.randint(0, 1000, (batch_size, max_length))
        attention_mask = torch.ones((batch_size, max_length))
        
        class TokenizerOutput:
            pass
        
        output = TokenizerOutput()
        output.input_ids = input_ids
        output.attention_mask = attention_mask
        return output

# Create a minimal dataset
dataset = [
    {
        'context': 'This is context 1',
        'response': 'This is response 1',
        'category': 'Toxicity Agreement',
        'label': 'Unsafe'
    },
    {
        'context': 'This is context 2',
        'response': 'This is response 2',
        'category': 'Unauthorized Expertise',
        'label': 'Safe'
    },
    {
        'context': 'This is context 3',
        'response': 'This is response 3',
        'category': 'Offending User',
        'label': 'Unsafe'
    },
    {
        'context': 'This is context 4',
        'response': 'This is response 4',
        'category': 'Biased Opinion',
        'label': 'Safe'
    }
]

tokenizer = MinimalTokenizer()

ARGS = {
    'dataset': dataset,
    'tokenizer': tokenizer,
    'batchsize': 2,
    'padsize': 128,
    'want_label': 1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm, transformers — 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
evaluate Not yet run
unverified (harvested, no recorded run)
thu-coai/diasafety/codes/mix_train.py
code served (permissive licence) · get_code("c2357b08a012370c")
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, 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 3 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

Dialogue safety problems severely limit the real-world deployment of neural conversational models and have attracted great research interests recently. However, dialogue safety problems remain under-defined and the corresponding dataset is scarce. We propose a taxonomy for dialogue safety specifically designed to capture unsafe behaviors in human-bot dialogue settings, with focuses on context-sensitive unsafety, which is under-explored in prior works. To spur research in this direction, we compile DiaSafety, a dataset with rich context-sensitive unsafe examples. Experiments show that existing safety guarding tools fail severely on our dataset. As a remedy, we train a dialogue safety classifier to provide a strong baseline for context-sensitive dialogue unsafety detection. With our classifier, we perform safety evaluations on popular conversational models and show that existing dialogue systems still exhibit concerning context-sensitive safety problems.

For agents

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

get_harvested_code_for_paper("2110.08466")
get_code_for_paper("2110.08466")
have("2110.08466")

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

curl https://syntology.ai/api/ran/2110.08466.json

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

[![Syntology run record](https://syntology.ai/api/ran/2110.08466.svg)](https://syntology.ai/paper/2110.08466)

Connect an agent — have() is free.