SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2305.02556 · ACL · 2023

Faithful Question Answering with Monte-Carlo Planning

Changshui Zhang, Hongming Zhang, Dong Yu, Ruixin Hong, Hong Zhao

arXiv · PDF · Open in the Atlas

Code that ran

3 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (3 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 3 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/2305.02556.json · badge: /api/ran/2305.02556.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
Raising-hrx/FAME marked official in Papers with Code
Papers with Code archive
3 ran · 1 no recorded run
FunctionStatusWhere it lives
MCTS_Node Ran
ran on a synthesized input · constructed an object; no result was computed
Raising-hrx/FAME/code/RL_agent.py
code served (permissive licence) · get_code("04f821bc4c441568")
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 an instance of its own class, MCTS_Node: the run constructed the object and did not call it, so no result was computed. 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

class MockEnv:
    """Mock environment that supports copy_env() method"""
    def __init__(self):
        self.state = "initial_state"
    
    def copy_env(self):
        new_env = MockEnv()
        new_env.state = self.state
        return new_env

# Set random seed for reproducibility
np.random.seed(42)

# Create a mock node_env
node_env = MockEnv()

# Create prior value
prior = 0.75

# Create node_action (optional, can be None for root node)
node_action = None

ARGS = {
    "node_env": node_env,
    "prior": prior,
    "node_action": node_action,
}
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 MCTS_Node — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a MCTS_Node 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
MCTS_Searcher Ran
ran on a synthesized input · constructed an object; no result was computed
Raising-hrx/FAME/code/RL_agent.py
code served (permissive licence) · get_code("5f5950dea6862b90")
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 an instance of its own class, MCTS_Searcher: the run constructed the object and did not call it, so no result was computed. 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 torch
import torch.nn as nn

class MockEnv:
    """Mock environment that supports copy_env()"""
    def __init__(self):
        self.state = "initial_state"
        self.step_count = 0
    
    def copy_env(self):
        new_env = MockEnv()
        new_env.state = self.state
        new_env.step_count = self.step_count
        return new_env

class MockController(nn.Module):
    """Mock controller that generates actions"""
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 8)
    
    def forward(self, state):
        # Returns mock action scores
        return torch.randn(2, 4)
    
    def generate(self, state_str, **kwargs):
        # Mock generation of candidate actions
        return [
            {
                'type': 'retrieve',
                'query_id': 'query_1',
                'controller_generate_score': 0.8
            },
            {
                'type': 'reason',
                'step': {
                    'pre_id': ['step_1', 'step_2'],
                    'con_sent': 'conclusion sentence'
                },
                'controller_generate_score': 0.6,
                'use_module': False
            }
        ]

class MockLinearizer:
    """Mock state linearizer"""
    def __call__(self, env):
        return f"state_{env.step_count}"

class MockConstraints:
    """Mock constraints checker"""
    def __call__(self, actions, env):
        return actions

ARGS = {
    'interact_args': {
        'linearize_state_form': MockLinearizer(),
        'controller_generate_args': {
            'max_length': 128,
            'num_beams': 3,
            'temperature': 0.7
        },
        'constraints_args_reason': MockConstraints(),
        'force_reason_strategy': None,
        'puct': 1.25,
        'discount': 0.99
    }
}
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 MCTS_Searcher — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a MCTS_Searcher 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
MinMaxStats Ran
ran on a synthesized input · constructed an object; no result was computed
Raising-hrx/FAME/code/RL_agent.py
code served (permissive licence) · get_code("5fb7146907888b93")
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 an instance of its own class, MinMaxStats: the run constructed the object and did not call it, so no result was computed. 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)

ARGS = {}
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 MinMaxStats — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a MinMaxStats 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
Action Not yet run
unverified (harvested, no recorded run)
Raising-hrx/FAME/code/RL_agent.py
code served (permissive licence) · get_code("913a8ad74b5da2f4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 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.

FunctionAlso present inEarliest-datedOthers named
Action
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2401.13246 (2024-01)
SEER: Facilitating Structured Reasoning and Explanation via Reinforcement Learning

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

Although large language models demonstrate remarkable question-answering performances, revealing the intermediate reasoning steps that the models faithfully follow remains challenging. In this paper, we propose FAME (FAithful question answering with MontE-carlo planning) to answer questions based on faithful reasoning steps. The reasoning steps are organized as a structured entailment tree, which shows how premises are used to produce intermediate conclusions that can prove the correctness of the answer. We formulate the task as a discrete decision-making problem and solve it through the interaction of a reasoning environment and a controller. The environment is modular and contains several basic task-oriented modules, while the controller proposes actions to assemble the modules. Since the search space could be large, we introduce a Monte-Carlo planning algorithm to do a look-ahead search and select actions that will eventually lead to highquality steps. FAME achieves advanced performance on the standard benchmark. It can produce valid and faithful reasoning steps compared with large language models with a much smaller model size.

For agents

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

get_harvested_code_for_paper("2305.02556")
get_code_for_paper("2305.02556")
have("2305.02556")

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

curl https://syntology.ai/api/ran/2305.02556.json

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

[![Syntology run record](https://syntology.ai/api/ran/2305.02556.svg)](https://syntology.ai/paper/2305.02556)

Connect an agent — have() is free.