FlowingFlowing
← Back to all development cases
Development case · blind model judgment pending independent human review; not a formal Benchmark conclusion.
AutocompleteEvidence version preferred137 / 162 · 4168fe7f868b2c44

Brain-Inspired Machine Intelligence: A Survey of Neurobiologically-Plausible Credit Assignment

Psychology and cognitive neuroscience · 2312.09257v2

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

Brain-Inspired Machine Intelligence: A Survey of Neurobiologically-Plausible Credit Assignment

3 Discussion

…layer effects and neural competition patterns [ 319 , 313 ] . ; all of which can be locally adjusted with the same goodness functional ℒ ℓ ​ ( 𝐳 ℓ , c ) \mathcal{L}^{\ell}(\mathbf{z}^{\ell},c) . Other self-supervised forward-only schemes either design local predictive contrastive losses [ 185 ] (which further mixes a local prediction of neural activity to produce a modulatory signal per neuron) or mutual information preservation [ 249 , 253 , 465 ] functionals (with modern inspirations such as [ 404 , 104 ] , which mixes local backprop with decoupled, block-local self-supervised functionals).

…mputing gradients in accordance with forward-mode differentiation [ 460 , 189 , 424 , 293 , 357 ] . Finally, it is important to point out that forward-only algorithms, notably signal propagation and the recurrent form FF/PFF, have been generalized to operate in the context of spiking neural networks [ 238 , 204 , 313 ] . This marks the beginning of an important move to investigation of such biological credit assignment schemes in the context of energy-efficient hardware; for example, an instantiation of FF learning has been demonstrated to operate well on an optical (fiber) platform [ 305 ] .

↓ 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

Forward-only and local learning rules have likewise been extended to spiking neural networks, where the sparsity of synaptic spiking inputs and the event-driven nature of neural processing can be leveraged by energy-efficient hardware implementations. This positions such schemes as a bridge between biological plausibility and the energy reductions that spiking substrates can offer relative to conventional artificial neural networks.

Passages supplied to the Evidence version

An Introduction to Probabilistic Spiking Neural Networks: Probabilistic Models, Learning Rules, and Applications

The sparsity of the synaptic spiking inputs and the corresponding event-driven nature of neural processing can be leveraged by energy-efficient hardware implementations, which can offer significant energy reductions as compared to conventional artificial neural networks (ANNs). The design of trainin…

Read full passage excerpt

The sparsity of the synaptic spiking inputs and the corresponding event-driven nature of neural processing can be leveraged by energy-efficient hardware implementations, which can offer significant energy reductions as compared to conventional artificial neural networks (ANNs). The design of training algorithms lags behind the hardware implementations. Most existing training algorithms for SNNs have been designed either for biological plausibility or through conversion from pretrained ANNs via rate encoding. This article provides an introduction to SNNs by focusing on a probabilistic signal processing methodology that enables the direct derivation of learning rules by leveraging the unique time-encoding capabilities of SNNs.

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

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, grounded continuation that bridges from the draft's discussion of forward-only algorithms in spiking neural networks to the energy-efficient hardware implications mentioned in the source. It accurately draws from the source's key points about sparsity of synaptic spiking inputs, event-driven processing, and energy-efficient hardware implementations offering reductions relative to conventional ANNs. The paragraph connects logically to the preceding text about SNN generalizations and the optical platform demonstration, maintaining academic tone and scope. Output B is empty despite the clear task requirement for content continuation, making it unusable.

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.