SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2412.17596 · 2024

Evaluating LLMs' Divergent Thinking Capabilities for Scientific Idea Generation with Minimal Context

arXiv · PDF · Open in the Atlas

Code that ran

4 of 11 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 4 have no contract verdict; 0 failures of our instrument; 7 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/2412.17596.json · badge: /api/ran/2412.17596.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
x66ccff/liveideabench canonical (classifier label)
classifier basis not recorded on the link
4 ran · 7 no recorded run
FunctionStatusWhere it lives
extract_final_idea Ran
ran on a synthesized input
x66ccff/liveideabench/utils/LLM.py
code served (permissive licence) · get_code("6ce5ba666cb8419a")
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 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
import torch.nn as nn

torch.manual_seed(0)

# Construct arguments for extract_final_idea
text_with_marker = """
Some initial discussion here.
**Final Idea:** This is the final idea that should be extracted.
Some more text after.
"""

ARGS = {
    "text": text_with_marker
}
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 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
get_values_from_dict Ran
ran on a synthesized input
x66ccff/liveideabench/utils/LLM.py
code served (permissive licence) · get_code("7884528ec3972cb4")
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 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 json
import re

# The function get_values_from_dict takes a single parameter: text (str)
# It extracts JSON formatted scores from text and returns a list of values

# Test case 1: JSON block with markdown formatting
text1 = """
Here are the evaluation scores:
json
{"originality": 8, "feasibility": 7, "clarity": 9}

"""

# Test case 2: Plain braces without markdown
text2 = """
The ratings are: {"originality": 5, "feasibility": 6, "clarity": 7}
"""

# Test case 3: Key-value pairs with quotes
text3 = '''
Evaluation results:
"originality": 4
"feasibility": 5
"clarity": 6
'''

# Test case 4: Key-value pairs without quotes
text4 = """
Score breakdown:
originality: 9
feasibility: 8
clarity: 7
"""

# Test case 5: No valid scores
text5 = "This text has no scoring information at all."

# We'll use test case 1 as the main example
ARGS = {
    "text": text1
}
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 int — 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_response_rejected Ran
ran on a synthesized input
x66ccff/liveideabench/utils/LLM.py
code served (permissive licence) · get_code("1ae45645f4e87838")
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

ARGS = {
    "text": "I'm sorry, I cannot help with that request as an AI assistant."
}
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
stable_hash Ran
ran on a synthesized input
x66ccff/liveideabench/utils/utils.py
code served (permissive licence) · get_code("7cf258cb8bc200e5")
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 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 hashlib

ARGS = {
    "text": "hello world"
}
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 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
calculate_embedding_similarity Not yet run
unverified (harvested, no recorded run)
x66ccff/liveideabench/hash.py
code served (permissive licence) · get_code("d1d3e83cca583daf")
How it ran

no invocation recorded.

calculate_jaccard_similarity Not yet run
unverified (harvested, no recorded run)
x66ccff/liveideabench/hash.py
code served (permissive licence) · get_code("c272689c24a64d2f")
How it ran

no invocation recorded.

calculate_tfidf_similarity Not yet run
unverified (harvested, no recorded run)
x66ccff/liveideabench/hash.py
code served (permissive licence) · get_code("bfb06d024ab1b947")
How it ran

no invocation recorded.

check_duplicate_entries Not yet run
unverified (harvested, no recorded run)
x66ccff/liveideabench/utils/database.py
code served (permissive licence) · get_code("44f0dd02f075cdc7")
How it ran

no invocation recorded.

query_results Not yet run
unverified (harvested, no recorded run)
x66ccff/liveideabench/utils/database.py
code served (permissive licence) · get_code("16494324701c93e6")
How it ran

no invocation recorded.

run_with_timeout Not yet run
unverified (harvested, no recorded run)
x66ccff/liveideabench/utils/utils.py
code served (permissive licence) · get_code("72563baa4eb2721e")
How it ran

no invocation recorded.

save_result Not yet run
unverified (harvested, no recorded run)
x66ccff/liveideabench/utils/database.py
code served (permissive licence) · get_code("8344c18f6c6bbafd")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 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 11 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

While Large Language Models (LLMs) demonstrate remarkable capabilities in scientific tasks such as literature analysis and experimental design (e.g., accurately extracting key findings from papers or generating coherent experimental procedures), existing evaluation benchmarks primarily assess performance using rich contextual inputs. We introduce LiveIdeaBench, a comprehensive benchmark evaluating LLMs' scientific idea generation by assessing divergent thinking capabilities using single-keyword prompts. Drawing from Guilford's creativity theory, our benchmark employs a dynamic panel of state-of-the-art LLMs to assess generated ideas across five key dimensions: originality, feasibility, fluency, flexibility, and clarity. Through extensive experimentation with over 40 leading models across 1,180 keywords spanning 22 scientific domains, we reveal that the scientific idea generation capabilities measured by our benchmark, are poorly predicted by standard metrics of general intelligence. Our results demonstrate that models like QwQ-32B-preview achieve creative performance comparable to top-tier models such as claude-3.7-sonnet:thinking, despite significant gaps in their general intelligence scores. These findings highlight the need for specialized evaluation benchmarks for scientific idea generation and suggest that enhancing these idea generation capabilities in LLMs may require different training strategies than those used for improving general problem-solving abilities, potentially enabling a wider range of AI tools tailored for different stages of the scientific process.

For agents

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

get_harvested_code_for_paper("2412.17596")
get_code_for_paper("2412.17596")
have("2412.17596")

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

curl https://syntology.ai/api/ran/2412.17596.json

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

[![Syntology run record](https://syntology.ai/api/ran/2412.17596.svg)](https://syntology.ai/paper/2412.17596)

Connect an agent — have() is free.