SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2406.12084 · EMNLP · 2024

When Reasoning Meets Information Aggregation: A Case Study with Sports Narratives

Kaiqiang Song, Fei Liu, Dong Yu, Sangwoo Cho, Wenlin Yao, Xiaoyang Wang, Yebowen Hu, Hassan Foroosh

arXiv · PDF · Open in the Atlas

Code that ran

14 of 15 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; 2 honoured the contract we drafted, 2 violated it and 0 have no contract verdict; 10 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/2406.12084.json · badge: /api/ran/2406.12084.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
yebowenhu/sportsgen canonical (classifier label)
classifier basis: census
14 ran · 1 no recorded run
FunctionStatusWhere it lives
DCA Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yebowenhu/sportsgen/utils/metric.py
code served (permissive licence) · get_code("46f0ccba5654e9e5")
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 float64, ndim 0, 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 numpy as np

# Set up valid arguments for the DCA function
pred_list = [1.2, 2.1, 3.5, 4.8]
tgt_list = [1.0, 2.0, 3.0, 5.0]
bucket_num = 5

ARGS = {
    "pred_list": pred_list,
    "tgt_list": tgt_list,
    "bucket_num": bucket_num
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
batchit Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/benchmark.py
code served (permissive licence) · get_code("890a7860ac2196ac")
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. 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
import numpy as np

torch.manual_seed(0)

# Create a simple corpus - the function iterates over it and groups rows into batches
# The corpus should be iterable and contain rows (can be strings, numbers, tensors, etc.)
corpus = [
    "sample text 1",
    "sample text 2",
    "sample text 3",
    "sample text 4",
    "sample text 5",
    torch.randn(4, 3),
    [1, 2, 3],
    np.array([4, 5, 6]),
    {"key": "value"},
    42,
]

# Batch size parameter
size = 3

ARGS = {
    "corpus": corpus,
    "size": size,
}
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 list — a list 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
build_tree_with_probabilities Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/utils/GameGenerator.py
code served (permissive licence) · get_code("863c3519e1b9c82e")
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 DiGraph. 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 networkx as nx

# Construct valid arguments for build_tree_with_probabilities
# The function takes paths: a list of sequences where each sequence represents a path
# It builds a directed graph with edge probabilities based on path frequencies

ARGS = {
    "paths": [
        ["A", "B", "C", "D"],
        ["A", "B", "E"],
        ["A", "C", "D"],
        ["B", "C", "F"],
        ["B", "C", "D"],
    ]
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
networkx — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type DiGraph — a DiGraph output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a DiGraph output the recorded ndim 1 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
discount_factors Ran
ran, honoured the contract we drafted
yebowenhu/sportsgen/utils/metric.py
code served (permissive licence) · get_code("01005644f05aa98b")
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 numpy as np

# The function takes a single parameter 'buckets' which should be an integer
# The function uses it to create an arange and calculate discount factors
# buckets is used in the arithmetic: np.arange(0, 1, 1 / (buckets + 1))

ARGS = {
    "buckets": 5
}
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 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
is_number Ran
ran, violated the contract we drafted
yebowenhu/sportsgen/utils/stats.py
code served (permissive licence) · get_code("316380478bf0bc96")
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 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
import numpy as np

torch.manual_seed(0)

# The function is_number takes input_value and tries to convert it to float
# It returns True if conversion succeeds, False if ValueError is raised

# Valid test cases for input_value:
# - integers: 5, -10
# - floats: 3.14, -2.5
# - strings that represent numbers: "42", "3.14", "-5"
# - strings that don't represent numbers: "hello"
# - numpy scalars and tensors

ARGS = {
    "input_value": 42  # integer that can be converted to float
}
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 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
is_player_in_team Ran
ran, violated the contract we drafted
yebowenhu/sportsgen/utils/NBAPlayer.py
code served (permissive licence) · get_code("88dea365876e43d4")
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 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)

player_name = {
    'player': 'Alice'
}

team = [
    {'name': 'Alice'},
    {'name': 'Bob'},
    {'name': 'Charlie'}
]

ARGS = {
    'player_name': player_name,
    'team': team
}
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 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
load_buckets Ran
ran, honoured the contract we drafted
yebowenhu/sportsgen/utils/metric.py
code served (permissive licence) · get_code("09949dd0641062e9")
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, ndim 2, dtype family int; 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 torch
import numpy as np

# Analyze the function:
# - total: an integer used in range and arithmetic operations
# - buckets: an integer used in arithmetic and comparisons

ARGS = {
    "total": 100,
    "buckets": 5
}
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 tuple, ndim 2, dtype family int — a tuple 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
load_folder Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/utils/stats.py
code served (permissive licence) · get_code("ecfc1a8bca988044")
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 os
import tempfile
import json

# Create a temporary directory with some .json files
temp_dir = tempfile.mkdtemp()

# Create a few dummy .json files
json_file_1 = os.path.join(temp_dir, "file1.json")
json_file_2 = os.path.join(temp_dir, "file2.json")
json_file_3 = os.path.join(temp_dir, "file3.json")

with open(json_file_1, 'w') as f:
    json.dump({"data": "test1"}, f)

with open(json_file_2, 'w') as f:
    json.dump({"data": "test2"}, f)

with open(json_file_3, 'w') as f:
    json.dump({"data": "test3"}, f)

# Also create a non-.json file to ensure it's not picked up
non_json_file = os.path.join(temp_dir, "file.txt")
with open(non_json_file, 'w') as f:
    f.write("not json")

ARGS = {
    "folder": 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 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
load_json Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/benchmark.py
code served (permissive licence) · get_code("40c9c445b5243b03")
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 json
import tempfile
import os

# Create a temporary JSON file with test data
test_data = {
    "key1": "value1",
    "key2": [1, 2, 3],
    "key3": {"nested": "object"}
}

# Create a temporary directory and file
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.json")

with open(temp_file, 'w') as f:
    json.dump(test_data, f)

ARGS = {
    "fpath": 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 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
load_json Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/utils/GameGenerator.py
code served (permissive licence) · get_code("3fb0a38707708bcc")
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 json
import tempfile
import os

# Create a temporary JSON file with valid content
temp_dir = tempfile.mkdtemp()
json_file_path = os.path.join(temp_dir, "test.json")

test_data = {
    "key1": "value1",
    "key2": 42,
    "key3": [1, 2, 3],
    "key4": {"nested": "object"}
}

with open(json_file_path, 'w') as f:
    json.dump(test_data, f)

ARGS = {
    "file_path": json_file_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 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
load_pickle Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/utils/GameGenerator.py
code served (permissive licence) · get_code("045e59a92bfe5692")
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 pickle
import tempfile
import os

# Create a temporary pickle file
temp_dir = tempfile.mkdtemp()
test_data = {"key": "value", "number": 42, "list": [1, 2, 3]}
temp_file = os.path.join(temp_dir, "test.pkl")

with open(temp_file, 'wb') as f:
    pickle.dump(test_data, f)

ARGS = {
    "file_path": 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 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
load_quarters Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/utils/stats.py
code served (permissive licence) · get_code("3cd34451dc8e1b47")
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 temporary JSON files that the function will read
temp_dir = tempfile.mkdtemp()

# Create first test file
file1_path = os.path.join(temp_dir, "game1.json")
file1_data = {
    "pbp": [
        {"event": 1, "quarter": 1},
        {"event": 2, "quarter": 1},
        {"event": 3, "quarter": 2},
        {"event": 4, "quarter": 2},
        {"event": 5, "quarter": 3},
        {"event": 6, "quarter": 3}
    ]
}
with open(file1_path, 'w') as f:
    json.dump(file1_data, f)

# Create second test file
file2_path = os.path.join(temp_dir, "game2.json")
file2_data = {
    "pbp": [
        {"event": 7, "quarter": 1},
        {"event": 8, "quarter": 1},
        {"event": 9, "quarter": 2},
        {"event": 10, "quarter": 2},
        {"event": 11, "quarter": 3},
        {"event": 12, "quarter": 3}
    ]
}
with open(file2_path, 'w') as f:
    json.dump(file2_data, f)

# Define ARGS with the file_list parameter
ARGS = {
    "file_list": [file1_path, file2_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
load_team_profile Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/utils/NBAPlayer.py
code served (permissive licence) · get_code("968d3fe53c0d7d21")
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
import numpy as np

# Construct valid arguments for load_team_profile
# The function expects a dict with 'strong_team' and 'weak_team' keys
# Each containing a list of player dicts with specific structure

def create_player(name, position, inside_score, outside_score):
    """Helper to create a player dict with required structure"""
    return {
        'name': name,
        'position': position,
        'general_abilities': {
            'Inside Scoring': inside_score,
            'Outside Scoring': outside_score
        }
    }

# Create strong team with 5+ players covering different positions
strong_team = [
    create_player('Player1', 'PG', 80, 85),
    create_player('Player2', 'SG', 82, 88),
    create_player('Player3', 'SF', 78, 80),
    create_player('Player4', 'PF', 85, 75),
    create_player('Player5', 'C', 88, 70),
    create_player('Player6', 'PG/SG', 75, 76),
]

# Create weak team with 5+ players covering different positions
weak_team = [
    create_player('Player7', 'PG', 60, 65),
    create_player('Player8', 'SG', 62, 68),
    create_player('Player9', 'SF', 58, 60),
    create_player('Player10', 'PF', 65, 55),
    create_player('Player11', 'C', 68, 50),
    create_player('Player12', 'SF/PF', 55, 56),
]

team_data_obj = {
    'strong_team': strong_team,
    'weak_team': weak_team
}

ARGS = {
    'team_data_obj': team_data_obj
}
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 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
team_score Ran
ran; our contract draft was wrong, not the code
yebowenhu/sportsgen/benchmark.py
code served (permissive licence) · get_code("ce4e3e460565736b")
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

# Construct a simple team_players dictionary
team_players = {
    "Team_A": ["Player1", "Player2", "Player3"],
    "Team_B": ["Player4", "Player5", "Player6"],
    "Team_C": ["Player7", "Player8"]
}

ARGS = {
    "team_players": team_players
}
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
select_team_players Not yet run
unverified (harvested, no recorded run)
yebowenhu/sportsgen/utils/NBAPlayer.py
code served (permissive licence) · get_code("2e9cd845f06720f7")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 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 15 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

Reasoning is most powerful when an LLM accurately aggregates relevant information. We examine the critical role of information aggregation in reasoning by requiring the LLM to analyze sports narratives. To succeed at this task, an LLM must infer points from actions, identify related entities, attribute points accurately to players and teams, and compile key statistics to draw conclusions. We conduct comprehensive experiments with real NBA basketball data and present SPORTSGEN, a new method to synthesize game narratives. By synthesizing data, we can rigorously evaluate LLMs' reasoning capabilities under complex scenarios with varying narrative lengths and density of information. Our findings show that most models, including GPT-4o, often fail to accurately aggregate basketball scores due to frequent scoring patterns. Open-source models like Llama-3 further suffer from significant score hallucinations. Finally, the effectiveness of reasoning is influenced by narrative complexity, information density, and domain-specific terms, highlighting the challenges in analytical reasoning tasks. 1 * Work done during Yebowen Hu's internship; Kaiqiang Song and Sangwoo Cho were full-time researchers at Tencent AI Lab, Seattle, USA at the time of this work.

For agents

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

get_harvested_code_for_paper("2406.12084")
get_code_for_paper("2406.12084")
have("2406.12084")

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

curl https://syntology.ai/api/ran/2406.12084.json

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

[![Syntology run record](https://syntology.ai/api/ran/2406.12084.svg)](https://syntology.ai/paper/2406.12084)

Connect an agent — have() is free.