FlowingFlowing
← Back to all development cases
Development case · blind model judgment pending independent human review; not a formal Benchmark conclusion.
AutocompleteWithout cards better145 / 162 · f411fb1f5c6804f6

A Survey on Improving Human Robot Collaboration through Vision-and-Language Navigation

Psychology and cognitive neuroscience · 2512.00027v1

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 positions and cards selected?

The evaluation uses seven fields with forty fixed ordinary autocomplete positions in each, for 280 total. A position does not need to require a literature citation in its next sentence. Each Flowing request receives three to five retrieved cards: five when all pass the relevance gate, without padding with weak cards. The run supplied 1,355 cards: 244 positions received five, 27 received four, and nine received three.

ADOPTION · 150/280

Adoption coverage

After drafting, the model returns an evidenceUse array identifying cards that materially affected facts, correction, direction, scope, terminology, or expression. A nonempty array counts as adoption.

OUTCOMES · 70 / 7 / 30

Full outcomes after adoption

Paired outcomes are counted only when the model reports adoption and both arms produce text. Among 107 comparable positions, Flowing was preferred in 70, 7 tied, and no-Evidence was preferred in 30; one tie was unusable in both arms.

How are the two table percentages calculated?

ADOPTION COVERAGE

150adopted
÷
280all positions
=
54%

FLOWING PREFERENCE AFTER ADOPTION

70Flowing preferred
÷
107comparable adopted positions
=
65%

evidenceUse is the model’s structured report of its own generation process. It does not require a citation and is not rejected merely because a faithful source paraphrase is not verbatim. Actual manuscript citations continue to use separate strict source verification. The previous 140-position result remains retained for traceability.

These are development-stage model-reported adoption and blind-judge results 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 Survey on Improving Human Robot Collaboration through Vision-and-Language Navigation

1. INTRODUCTION > 1.1. Our Contributions

…nts to interpret a single instruction before navigating, while others support free-form dialogue with an oracle. Similarly, task objectives range from precise route following to dynamic exploration and object interaction. Even seemingly simple directives such as “Turn left, climb the stairs, enter the bathroom” pose challenges for computational agents. They must decompose such instructions into sub-goals, ground each step in real-world objects and dynamics, recognize visual cues (e.g., identifying the bathroom), and execute actions accurately while knowing when to stop ( Jain et al., 2019 ) .

…ipulation, and visual communication in collaborative tasks. Furthermore, reviews of collaborative robotics emphasize safety, control performance, and intuitive interaction in industrial applications. Our paper examines existing works, challenges, and opportunities in developing robust collaborative VLN models that efficiently coordinate multiple robots while addressing the needs and preferences of human collaborators. The interdisciplinary nature of this field, as highlighted in ( Buxbaum et al., 2020 ) , underscores the need for a structured approach to guide future research and development.

↓ 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: Without cards better

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

Not preferred in this comparison

No continuation text was produced.

The model requested more evidence and did not continue.

Passages supplied to the Evidence version

Vision-and-Language Navigation: A Survey of Tasks, Methods, and Future Directions

Multi-agent VLN tasks require development in swarm intelligence, information communication, and performance evaluation. MeetUp! ( Ilinykh et al., 2019 ) is a two-player coordination game where players move in a visual environment to find each other. VLN studies the relationship between the human and…

Read full passage excerpt

Multi-agent VLN tasks require development in swarm intelligence, information communication, and performance evaluation. MeetUp! ( Ilinykh et al., 2019 ) is a two-player coordination game where players move in a visual environment to find each other. VLN studies the relationship between the human and the environment in Figure 1 , yet here humans are oracles simply observing (but not acting on) the environment. Collaboration between humans and robots is crucial for them to work together as teams (e.g., as personal assistants or helping in construction). Future work may target at collaborative VLN between multiple agents or between human and agents.

Model-reported card use

The model did not report that any card materially affected this completion.

No Evidence

Same model and position · no retrieved passages · first output

Preferred by blind review

Existing surveys have largely examined VLN as a single-agent problem, focusing on instruction grounding and navigation performance rather than the coordination demands that arise when robots must work alongside humans. This leaves open how collaborative VLN models should be structured to support shared goals, role allocation, and communication with human partners.

Result of this comparison

The blinded model review preferred the no-Evidence version. Its original rationale appears below.

Read the original model judgment

Output A provides a coherent, grounded paragraph that bridges the gap between the preceding discussion of single-agent VLN challenges and the subsequent focus on collaborative robotics. It accurately notes that existing surveys have focused on single-agent VLN (consistent with the source's statement that 'VLN studies the relationship between the human and the environment...yet here humans are oracles simply observing'), and identifies the open question of how collaborative VLN models should handle coordination, shared goals, and human-robot communication—directly echoing the source's call for 'future work may target at collaborative VLN between multiple agents or between human and agents.' The paragraph flows naturally from the draft's discussion of instruction following challenges to the paper's stated focus on collaborative VLN. Output B returns an empty text with 'needs_evidence' action, but the task clearly requires content and the source does provide relevant factual context about the gap between single-agent VLN and collaborative VLN. An empty response fails to complete the autocomplete task.

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.