SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2210.01240 · ICLR · 2023

Language Models Are Greedy Reasoners: A Systematic Formal Analysis of Chain-of-Thought

Abulhair Saparov, He He

arXiv · PDF · Open in the Atlas

Code that ran

4 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; 1 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 2 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/2210.01240.json · badge: /api/ran/2210.01240.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
asaparov/prontoqa canonical (classifier label)
classifier basis: classifier_2025
4 ran · 1 no recorded run
FunctionStatusWhere it lives
get_count Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
asaparov/prontoqa/analyze_results.py
code served (permissive licence) · get_code("1d11c551594f3298")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned int64, ndim 0, dtype family int, 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 numpy as np

result_array = np.array([10, 5, 8, 3, 12, 7])
index = 3

ARGS = {
    'result_array': result_array,
    'index': index
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int64, ndim 0, dtype family int, 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
lighten_color Ran
ran, honoured the contract we drafted
asaparov/prontoqa/analyze_results.py
code served (permissive licence) · get_code("94b21ef3a27e0a02")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned tuple, 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 matplotlib.colors as mc

# The function takes a color (either a named color or a color specification)
# and an amount parameter (default 0.5) to lighten the color

# We can use a named color from matplotlib's color names
color = 'blue'

# Or we could use a hex color
# color = '#FF5733'

# Or an RGB tuple
# color = (0.5, 0.2, 0.8)

amount = 0.5

ARGS = {
    'color': color,
    'amount': amount
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
matplotlib — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — 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
rindex Ran
ran; our contract draft was wrong, not the code
asaparov/prontoqa/run_experiment.py
code served (permissive licence) · get_code("63403bf63d6ded29")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned int, ndim 0, dtype family int, 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.

items = [1, 2, 3, 2, 4, 2, 5]
element = 2

ARGS = {
    'items': items,
    'element': element
}
Setup block drafted by
a language model, Claude Sonnet 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 int, ndim 0, dtype family int, 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
wilson_conf_interval Ran
ran, violated the contract we drafted
asaparov/prontoqa/analyze_results.py
code served (permissive licence) · get_code("f24ebcccb5b1d226")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned tuple, 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

# wilson_conf_interval computes a confidence interval for a binomial proportion
# p: observed proportion (between 0 and 1)
# n: sample size (positive integer)
# z: z-score for desired confidence level (default 1.96 for 95% confidence)

ARGS = {
    'p': 0.6,
    'n': 100,
    'z': 1.96
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — 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
is_antecedent_provable Not yet run
unverified (harvested, no recorded run)
asaparov/prontoqa/run_experiment.py
code served (permissive licence) · get_code("0408f2ab52da6612")
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, 5 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 5 later) and 0 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
lighten_color 6 other papers
1 dated earlier, 5 later
2001.04032 (2020-01)
POPCORN: Partially Observed Prediction COnstrained ReiNforcement Learning
2301.08517 (2023-01)
Cohere: Managing Differential Privacy in Large Scale Systems
2305.15269 (2023-05)
Testing the General Deductive Reasoning Capacity of Large Language Models Using OOD Examples
and 3 more not named here
get_count 1 other paper
1 later
2305.15269 (2023-05)
Testing the General Deductive Reasoning Capacity of Large Language Models Using OOD Examples
is_antecedent_provable 1 other paper
1 later
2305.15269 (2023-05)
Testing the General Deductive Reasoning Capacity of Large Language Models Using OOD Examples
rindex 1 other paper
1 later
2305.15269 (2023-05)
Testing the General Deductive Reasoning Capacity of Large Language Models Using OOD Examples
wilson_conf_interval 1 other paper
1 later
2305.15269 (2023-05)
Testing the General Deductive Reasoning Capacity of Large Language Models Using OOD Examples

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

Large language models (LLMs) have shown remarkable reasoning capabilities given chain-of-thought prompts (examples with intermediate reasoning steps). Existing benchmarks measure reasoning ability indirectly, by evaluating accuracy on downstream tasks such as mathematical reasoning. However, it is unclear how these models obtain the answers and whether they rely on simple heuristics rather than the generated chain-of-thought. To enable systematic exploration of the reasoning ability of LLMs, we present a new synthetic question-answering dataset called PRONTOQA, where each example is generated from a synthetic world model represented in first-order logic. This allows us to parse the generated chain-ofthought into symbolic proofs for formal analysis. Our analysis on INSTRUCTGPT and GPT-3 shows that LLMs are quite capable of making correct individual deduction steps, and so are generally capable of reasoning, even in fictional contexts. However, they have difficulty with proof planning: When multiple valid deduction steps are available, they are not able to systematically explore the different options.

For agents

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

get_harvested_code_for_paper("2210.01240")
get_code_for_paper("2210.01240")
have("2210.01240")

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

curl https://syntology.ai/api/ran/2210.01240.json

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

[![Syntology run record](https://syntology.ai/api/ran/2210.01240.svg)](https://syntology.ai/paper/2210.01240)

Connect an agent — have() is free.