FlowingFlowing
← Back to all development cases
Development case · blind model judgment pending independent human review; not a formal Benchmark conclusion.
AutocompleteBoth unusable65 / 162 · 50b470bd06c76f6e

A review of machine learning applications in wildfire science and management

Climate and environmental science · 2003.00646v2

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 machine learning applications in wildfire science and management

5 Discussion > 5.6 Implications for fire management

…n). The ML community, however, recognizes the problem of interpretability and work is underway to develop methods that allow for greater interpretability of ML methods, including methods for DL (see for example, McGovern et al. [2019] ) or model-agnostic approaches [ Ribeiro et al., 2016 ] . Runge et al. [2019] further argue that casual inference methods should be used in conjunction with predictive models to improve our understanding of physical systems. Finally, it is worth noting that assessing variable importance (see Sec. 4.3.4 ) for a given model can play a role in model interpretation.

…[ Brenowitz and Bretherton, 2018 ] . Other recent approaches use ML methods to determine the solutions to nonlinear partial differential equations Raissi and Karniadakis [2018] , Raissi et al. [2019] . Such methods could find future applications in improving fire behaviour prediction models based on computationally expensive physics-based fire simulators, in coupled fire-atmosphere models, or in smoke dispersion modeling. In any case, the applications of ML that we have outlined are meant for illustrative purposes and are not meant to represent an exhaustive list of all possible applications.

↓ 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: Both unusable

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

Judged unusable

The product produced no insertable continuation text.

The model explicitly returned a no-completion marker.

Passages supplied to the Evidence version

Data-driven discretization: a method for systematic coarse graining of partial differential equations ↗

Here we introduce data driven discretization , a method for automatically learning effective long-wavelength dynamics from actual solutions to the known underlying equations. We use a neural network to learn a discretization for the true spatial derivatives of partial differential equations. We demo…

Read full passage excerpt

Here we introduce data driven discretization , a method for automatically learning effective long-wavelength dynamics from actual solutions to the known underlying equations. We use a neural network to learn a discretization for the true spatial derivatives of partial differential equations. We demonstrate that this approach is remarkably accurate, allowing us to integrate in time a collection of nonlinear equations in one spatial dimension at resolutions 4-8x coarser than is possible with standard finite difference methods.

No Evidence

Same model and position · no retrieved passages · first output

Judged unusable

The product produced no insertable continuation text.

The model explicitly returned a no-completion marker.

Judgment and task check

Public counts mark both outputs unusable; an empty continuation cannot complete this task.

Read the original model judgment

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

Both outputs return empty text with action 'complete'. The task requires continuing the manuscript with one coherent academic paragraph in the blank space between the two provided draft segments. The blank paragraph needs content that bridges the discussion of ML interpretability and variable importance (end of first segment) with the discussion of ML for solving nonlinear PDEs and fire behavior modeling (start of second segment). An empty response fails to fulfill the core task requirement of generating a continuation paragraph. The source provided discusses data-driven discretization for coarse-graining PDEs, which could potentially connect to the PDE-solving methods mentioned in the second segment, but neither output attempts to use this or any other material to bridge the gap. Both outputs are empty when the task clearly needs content.

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.