FlowingFlowing
← Back to all development cases
Development case · blind model judgment pending independent human review; not a formal Benchmark conclusion.
AutocompleteEvidence version preferred4 / 162 · cebee9387c3b261c

Bridging from single to collective cell migration: A review of models and links to experiments

Quantitative biology · 2011.10873v1

FLOWING EVIDENCE BENCHMARK

How do we tell whether Evidence helps?

At the same writing position, with the same model and task, how does supplying retrieved paper passages change the first output? We compare matched versions and retain ties, unusable outputs and incomplete reviews.

SAME DRAFT · EVIDENCE ON OR OFF

01 · FIXED WRITING POSITION

MANUSCRIPT

Same writing position ▌

Matched autocomplete at the same draft position

02 · TWO MATCHED INPUTS

SHARED BY BOTH

Manuscript context, model, task and prompt

A · With Evidence

Retrieved paper passages supplied

B · Without Evidence

No retrieved passages supplied

LLM

Same model and version

A → first output

B → first output

Blind judge agent

First outputs are anonymized as X and Y

Continuation review: accuracy, fit to the writing task and usability
Returns: X preferred / tie / Y preferred / both unusable

Order check: X / Y → Y / X

HOW DOES BLIND JUDGING WORK?

① Anonymize both outputs
The judge sees the same draft and both first outputs without knowing which received Evidence.

② Compare and swap order
The judge applies task-specific criteria in X/Y and then Y/X order.

③ Review disagreements
A third pass resolves disagreements. Incomplete reviews remain in the denominator.

How are autocomplete positions stratified?

Before seeing generation outcomes, we check whether a retrieved passage contains a specific proposition that directly supports the next writing move. Those positions appear in the left opportunity group; the rest are ordinary positions on the right. We select a balanced sample from admitted papers in each field. The 50/50 split is experimental, not a measure of how often either type occurs in writing.

AUTOCOMPLETE · 110

Biology, statistics and astrophysics

55 positions on each side; statistics uses the ten-paper rerun.

AUTOCOMPLETE · 52

Psychology and climate science

26 positions on each side; psychology includes nine papers and climate science four.

How are the table percentages calculated?

Across five fields, 51 of 81 left-column positions preferred the Evidence version. Ties, pairs where both versions were unusable and incomplete reviews remain in the denominator.

51Evidence version preferred
÷
81All positions in this group
=
63%Evidence preference in this group

Source contribution is a separate review: 16 of 22 Evidence wins entered into source review directly used retrieved papers; another 29 wins await review.

These are development-stage model judgments pending independent human review. They are not formal Benchmark conclusions and do not, on their own, establish causality.

Manuscript writing position

Text excerpt · not a PDF page

Research manuscript · excerpt

Bridging from single to collective cell migration: A review of models and links to experiments

5 Discussion > 5.3 Model-experiment links

Biology is an experimental science at its core, and models that illuminate its mysteries must eventually meet and concur with biological evidence. Modern methods such as machine learning can help match a vast dataset with the best candidate model(s) as shown elegantly by [ 118 ] . At the same time, modelers, physicists, and biophysicists can make real contributions by using their craft and theories to unlock important facts that are not at all apparent otherwise [ 42 ] .

Several of our examples demonstrate that some experimental papers have explicitly addressed signaling pathways that mediate cell–cell communication in collective cell behavior. As pointed out eloquently by a reviewer of this paper, “experiments can only tell [us … what are] upstream and downstream regulators. We need mathematical models to incorporate this information with the intracellular networks to form a signaling network on a multicellular level to study how a group of cells processes signals collaboratively.”

↓ Next paragraph was generated here

The highlighted text is near the writing position. This is a layout of the evaluation input, not a PDF screenshot.

A matched comparison at the same writing positionJudgment: Evidence version preferred

Both arms use the same model, task and manuscript context. Only the retrieved passages differ. The first recorded output from each arm is shown below.

Evidence

Same model and position · retrieved passages supplied · first output

Preferred by blind review

The reviewer's point identifies a division of labor rather than a gap in either approach: experiments establish which regulators act upstream and downstream of a signal, while models are needed to embed that information in an intracellular network and ask how a group of cells processes signals collaboratively. Network control frameworks illustrate how such intracellular information can be used to predict intervention targets, as shown for a leukemia signaling network and for the network controlling helper T cell differentiation.

Passages supplied to the Evidence version

Cell fate reprogramming by control of intracellular network dynamics

Here we develop a novel network control framework that integrates the structural and functional information available for intracellular networks to predict control targets. Formulated in a logical dynamic scheme, our approach drives any initial state to the target state with 100% effectiveness and n…

Read full passage excerpt

Here we develop a novel network control framework that integrates the structural and functional information available for intracellular networks to predict control targets. Formulated in a logical dynamic scheme, our approach drives any initial state to the target state with 100% effectiveness and needs to be applied only transiently for the network to reach and stay in the desired state. We illustrate our method’s potential to find intervention targets for cancer treatment and cell differentiation by applying it to a leukemia signaling network and to the network controlling the differentiation of helper T cells. We find that the predicted control targets are effective in a broad dynamic framework. Moreover, several of the predicted interventions are supported by experiments.

No Evidence

Same model and position · no retrieved passages · first output

Not preferred in this comparison

No continuation text was produced.

Model action：complete

Blind model judgment

The blinded model review preferred the Evidence version. Its original rationale appears below.

Read the original model judgment

Output A provides a coherent, usable continuation that directly addresses the reviewer's quoted point about the complementary roles of experiments and models in studying collective cell signaling. It grounds its claims in the supplied source about network control frameworks for intracellular networks, specifically mentioning the leukemia signaling network and helper T cell differentiation examples from the source—matching entity, relation, and scope without overclaiming. The paragraph maintains academic tone and flows logically from the draft's discussion of experiments identifying upstream/downstream regulators to the need for models to integrate this information multicellularly. Output B is empty when the task clearly requires content continuation; a blank paragraph cannot be inserted as a coherent academic paragraph.

Source review confirmed that this continuation directly used specific, verifiable information from the retrieved paper.

The passages are input material; their presence does not mean an output used them or that they support every claim. Draft and source passages are excerpted; no simulated scores or PDF appear here.