FlowingFlowing
← Back to all development cases
Development case · blind model judgment pending independent human review; not a formal Benchmark conclusion.
AutocompleteBoth unusable131 / 162 · 66d1907f048ccd6d

Quantum Computational Complexity -- From Quantum Information to Black Holes and Back

Astrophysics · 2110.14672v1

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

Quantum Computational Complexity -- From Quantum Information to Black Holes and Back

9 Summary and Outlook

…different quantity, since one is not allowed to use all possible gates but only those that act on the radiation without acting on the interior. A different holographic conjecture was proposed for this restricted complexity , which involves the area of the maximum cross-section of the wormhole and of the minimal surface in the throat that connects it to the asymptotic region. This shows that there are probably different notions of complexity that can be useful for answering different questions about the quantum information-theoretic aspects of gravity, and there is still much to be understood.

…bservables: low-point correlation functions (especially out-of-time-order correlators), and entanglement entropy. Quantum computational complexity captures properties of the quantum state of a system that are more refined than those visible through these observables. This is why it is sensitive to the evolution of the microstates in the ensemble corresponding to a black hole. It is likely that it can also be used to give new insights into the mechanisms underlying the approach to equilibrium and thermalization, and possibly detect new types of phase transitions (see e.g., [ 150 , 29 , 30 ] ).

↓ 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 generated prose, but the product admission rule blocked it as needing evidence. That raw prose has not been independently reviewed here.

Passages supplied to the Evidence version

Entanglement growth during thermalization in holographic systems ↗

Various nonlocal observables have a unified description in terms of the area of extremal surfaces of different dimensions. In the large distance limit, the evolution of an extremal surface, and thus the corresponding boundary observable, is controlled by the geometry around and inside the event hori…

Read full passage excerpt

Various nonlocal observables have a unified description in terms of the area of extremal surfaces of different dimensions. In the large distance limit, the evolution of an extremal surface, and thus the corresponding boundary observable, is controlled by the geometry around and inside the event horizon of the black hole, allowing us to identify regimes of pre-local- equilibration quadratic growth, post-local-equilibration linear growth, a memory loss regime, and a saturation regime with behavior resembling those in phase transitions. We also discuss possible bounds on the maximal rate of entanglement growth in relativistic systems.

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 = No Evidence; Output B = Evidence.

Both outputs are empty and fail to provide the required autocomplete paragraph. Output A returns empty text with action 'complete', which is unusable because the task explicitly requires continuing the manuscript with one coherent academic paragraph. Output B returns empty text with action 'needs_evidence', which is also unusable for the same reason. The blank paragraph in the draft clearly needs content that bridges the discussion of complexity's sensitivity to black hole microstate evolution (ending the prior paragraph) to the next section. The available source discusses entanglement growth during thermalization with regimes of growth and saturation resembling phase transitions, which could potentially connect to the draft's mention of 'approach to equilibrium and thermalization' and 'new types of phase transitions.' However, neither output attempts to use this or produce any paragraph at all. An empty result is only usable when no supplied source directly supports a card, but here the task is autocomplete and the draft has a clear gap requiring continuation.

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.