FlowingFlowing
← Back to all development cases
Development case · blind model judgment pending independent human review; not a formal Benchmark conclusion.
AutocompleteEvidence version preferred129 / 162 · 614bd02e252d012b

A Review of BioTree Construction in the Context of Information Fusion: Priors, Methods, Applications and Trends

Quantitative biology · 2410.04815v2

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 there are 81 source-opportunity positions. The original blind review preferred Evidence in 51; a task check moved one empty Evidence continuation to both unusable, leaving 50 in public counts. In the ordinary group, another pair of empty outputs moved from no winner to both unusable. Original verdicts remain visible on case pages.

50Evidence version preferred
÷
81All positions in this group
=
62%Evidence preference in this group

Source contribution is a separate review: 16 of 21 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

A Review of BioTree Construction in the Context of Information Fusion: Priors, Methods, Applications and Trends

6 Classical BioTree Construction Methods > 6.5 Limitations of Traditional BioTree Construction Methods

The rise of multi-omics approaches introduces complex data integration demands that traditional methods struggle to address. These methods, often tailored for single sequence types, face difficulties in capturing the biological context of genomic, transcriptomic, and proteomic interrelationships. Advances in statistical models are gradually improving adaptability, but the challenges of scalability and dimensionality remain significant.

Predefined evolutionary models, such as substitution models , simplify phylogenetic analysis but may not fully reflect real evolutionary dynamics, where rates vary across lineages and selective pressures differ among genes. This dependency introduces biases that modern flexible models aim to address, allowing for more accurate evolutionary representations.

↓ 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

A further limitation concerns the mapping from genotype to phenotype: because many genotypes can map to the same phenotype, and the number of genotypes underlying a given phenotype is often strongly biased, traditional methods that treat sequence variation as a direct proxy for phenotypic difference may misrepresent the evolutionary signal they are intended to capture.

Passages supplied to the Evidence version

From genotypes to organisms: State-of-the-art and perspectives of a cornerstone in evolutionary dynamics ↗

Nevertheless, important progress in this direction has been made through the use of GP maps that can be computationally explored and, more recently, through the development of quantitative approaches to shared generic properties. Among the latter, one of the most striking properties is a strong bias…

Read full passage excerpt

Nevertheless, important progress in this direction has been made through the use of GP maps that can be computationally explored and, more recently, through the development of quantitative approaches to shared generic properties. Among the latter, one of the most striking properties is a strong bias in the number of genotypes mapping to a phenotype. [ 12 , 4 ] This begs the question: Can this bias, which often extends over many orders of magnitude, affect evolutionary outcomes? Indeed, phenotypic bias, among other non-trivial properties of GP maps, does severely affect not only our understanding of how variation arises through random mutations, but also any accurate representation —be it metaphorical or formal— of evolutionary dynamics at large.

No Evidence

Same model and position · no retrieved passages · first output

Not preferred in this comparison

The product produced no insertable continuation text.

The model explicitly returned a no-completion marker.

Judgment and task check

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

Read the original model judgment

In this rationale, Output A = Evidence; Output B = No Evidence.

Output A provides a coherent, source-grounded continuation that bridges from the draft's discussion of evolutionary model biases to the genotype-phenotype mapping problem described in the supplied source. The source explicitly discusses 'a strong bias in the number of genotypes mapping to a phenotype' and how 'phenotypic bias...does severely affect...any accurate representation...of evolutionary dynamics at large.' Output A accurately incorporates this concept, noting that traditional methods treating 'sequence variation as a direct proxy for phenotypic difference may misrepresent the evolutionary signal'—a logical extension of both the draft's theme (model limitations/biases) and the source's content. Output B is empty, which fails the task requirement of providing 'one coherent academic paragraph' for autocomplete; this is not a case where deliberate emptiness is justified, as the source provides directly relevant material.

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.