SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.02787 · 2021

BiToD: A Bilingual Multi-Domain Dataset For Task-Oriented Dialogue Modeling

arXiv · PDF · Open in the Atlas

Code that ran

5 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, 1 violated it and 0 have no contract verdict; 4 failures of our instrument; 0 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/2106.02787.json · badge: /api/ran/2106.02787.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
HLTCHKUST/BiToD canonical (classifier label)
classifier basis: classifier_2025
5 ran · 0 no recorded run
FunctionStatusWhere it lives
compute_lev_span Ran
ran; our contract draft was wrong, not the code
HLTCHKUST/BiToD/preprocess.py
code served (permissive licence) · get_code("16527c949b0dad20")
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) 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 copy

def compute_lev_span(previous_state, new_state, intent):
    Lev = f'<API> {intent}'
    if intent == 'chat':
        return '<API>'
    old_state = copy.deepcopy(previous_state)
    if intent not in old_state:
        old_state[intent] = {}
    for slot in new_state[intent]:
        if old_state[intent].get(slot) != new_state[intent].get(slot):
            relation = new_state[intent][slot]['relation']
            values = [str(value) for value in new_state[intent][slot]['value']]
            values = '<value> '.join(values)
            Lev += f'<slot> {slot}<relation> {relation}<value> {values}'
    for slot in old_state[intent]:
        if slot not in new_state[intent]:
            print(intent, old_state[intent][slot])
            Lev += f'<slot> {slot}<unknow>'
    return Lev


# Setup code
previous_state = {
    'restaurant': {
        'location': {'relation': 'equal', 'value': ['downtown']},
        'cuisine': {'relation': 'equal', 'value': ['italian']},
    }
}

new_state = {
    'restaurant': {
        'location': {'relation': 'equal', 'value': ['downtown']},
        'cuisine': {'relation': 'equal', 'value': ['french']},
        'price': {'relation': 'equal', 'value': ['expensive']},
    }
}

intent = 'restaurant'

ARGS = {
    'previous_state': previous_state,
    'new_state': new_state,
    'intent': intent,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
is_int Ran
ran, violated the contract we drafted
HLTCHKUST/BiToD/evaluate.py
code served (permissive licence) · get_code("c6ce652bfb099994")
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) and returned bool, 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)

ARGS = {
    "val": "42"
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type bool, 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
knowledge2span Ran
ran; our contract draft was wrong, not the code
HLTCHKUST/BiToD/preprocess.py
code served (permissive licence) · get_code("263abb1c666a9c10")
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) 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)

# Construct a valid knowledge dictionary argument
knowledge = {
    'domain1': {
        'slot1': ['value1', 'value2', 'value3'],
        'slot2': 'single_value',
        'type': 'ignored_field',
        'description': 'also_ignored'
    },
    'domain2': {
        'slot3': ['val_a', 'val_b'],
        'slot4': 'another_value',
        '类别': 'ignored_chinese',
        '描述': 'also_ignored_chinese'
    }
}

ARGS = {
    'knowledge': knowledge
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
span2dict Ran
ran; our contract draft was wrong, not the code
HLTCHKUST/BiToD/evaluate.py
code served (permissive licence) · get_code("fe6b1382a7dc0e63")
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) 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.

# Setup code for span2dict function

def is_int(val):
    try:
        num = int(val)
    except ValueError:
        return False
    return True

def span2dict(api_span, api_names):
    state = {}
    api_list = api_span.split('<')
    for word in api_list:
        if word.startswith('API'):
            api_name = word.replace('API> ', '')
            if api_name in api_names:
                state[api_name] = {}
            slot = None
            relation = None
            value = None
        elif word.startswith('slot'):
            if value:
                value[0] = int(value[0]) if is_int(value[0]) else value[0]
                try:
                    state[api_name][slot] = {'relation': relation, 'value': value}
                except:
                    pass
            slot = word.replace('slot> ', '')
        elif word.startswith('relation'):
            relation = word.replace('relation> ', '')
            value = []
        elif word.startswith('value'):
            v = word.replace('value> ', '')
            value.append(v)
    try:
        if value:
            value[0] = int(value[0]) if is_int(value[0]) else value[0]
            state[api_name][slot] = {'relation': relation, 'value': value}
    except:
        print(f'FAILED: api:{api_name}, slot:{slot}, relation:{relation}, value:{value}')
    return state

# Construct valid arguments
api_span = "<API> weather<slot> temperature<relation> equals<value> 72<slot> location<relation> is<value> New York"
api_names = ["weather", "forecast", "alerts"]

ARGS = {
    "api_span": api_span,
    "api_names": api_names
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 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
state2span Ran
ran; our contract draft was wrong, not the code
HLTCHKUST/BiToD/preprocess.py
code served (permissive licence) · get_code("99cd7c02bcd44503")
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) 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

# Construct valid arguments for state2span function

# state: a nested dict with intents as keys, containing slot information
state = {
    'book_hotel': {
        'location': {
            'relation': 'equals',
            'value': ['New York']
        },
        'date': {
            'relation': 'after',
            'value': ['2024-01-15']
        }
    },
    'search_flight': {
        'destination': {
            'relation': 'equals',
            'value': ['Paris', 'London']
        }
    }
}

# required_slots: a dict mapping intent names to lists of required slot names
required_slots = {
    'book_hotel': ['location', 'date'],
    'search_flight': ['destination']
}

ARGS = {
    'state': state,
    'required_slots': required_slots
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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

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

Task-oriented dialogue (ToD) benchmarks provide an important avenue to measure progress and develop better conversational agents. However, existing datasets for end-to-end ToD modeling are limited to a single language, hindering the development of robust end-to-end ToD systems for multilingual countries and regions. Here we introduce BiToD, the first bilingual multi-domain dataset for end-to-end task-oriented dialogue modeling. BiToD contains over 7k multi-domain dialogues (144k utterances) with a large and realistic bilingual knowledge base. It serves as an effective benchmark for evaluating bilingual ToD systems and cross-lingual transfer learning approaches. We provide state-of-the-art baselines under three evaluation settings (monolingual, bilingual, and cross-lingual). The analysis of our baselines in different settings highlights 1) the effectiveness of training a bilingual ToD system compared to two independent monolingual ToD systems, and 2) the potential of leveraging a bilingual knowledge base and cross-lingual transfer learning to improve the system performance under low resource condition.

For agents

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

get_harvested_code_for_paper("2106.02787")
get_code_for_paper("2106.02787")
have("2106.02787")

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

curl https://syntology.ai/api/ran/2106.02787.json

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

[![Syntology run record](https://syntology.ai/api/ran/2106.02787.svg)](https://syntology.ai/paper/2106.02787)

Connect an agent — have() is free.