Yila
Édition scientifique en 6 couches · Articles & projets · Zéro dérive des données

Humaniseur académique

Élimine 11 catégories de tics d’écriture IA dans vos articles, réponses aux relecteurs et projets de recherche (NSF, NIH) en alignant chaque affirmation sur les preuves et sur votre propre style — sans modifier aucun chiffre, formule ou citation.

Essayer sur un exemple réel :
  • 0 chiffre, équation ou citation modifié
  • Activation explicite uniquement (jamais déclenché en arrière-plan)
  • Livre le texte révisé + le journal d’audit en 6 couches
Anatomie comparative avant / après

Ce qu’une passe académique en 6 couches retire — et ce qu’elle protège

Les reformulateurs génériques dégradent la rigueur scientifique. Basculez entre articles et demandes de financement ci-dessous.

Empirical Paper Pass7 AI tells removed
Avant · Brouillon assisté par IA

In recent years, graph representation learning has attracted increasing attention.2.5 Formulaic opener In this paper, we propose a novel2.4 Novelty padding method and demonstrate2.1 Over-claim verb through extensive experiments2.3 Empty intensifier that it significantly outperforms2.1 Unquantified stat existing approaches, paving the way for a new paradigm.2.2 Hype To the best of our knowledge, this is the first work to delve into this crucial problem.L1 delve + 2.4 First-claim

Après · Yila Humaniseur académique

Existing graph methods for imbalanced node classification reweight the loss but ignore label noise. We add a noise-aware reweighting term and evaluate it on three citation benchmarks, where it improves minority-class F1 by 3–6 points over focal loss while leaving majority-class accuracy within 1 point.

0 chiffre ou formule altéré100 % des citations conservéesVerbes ajustés aux preuves
Synthèse d’audit : Cut the textbook opener, "novel", "extensive", unbacked "significantly", "paving the way", and "delve"; downgraded "demonstrate" to "evaluate" and anchored the gain to an exact comparator and range.
Fiche mémo avant soumission

Les 11 tics d’écriture académiques IA repérés par les relecteurs

Utilisez cette grille de relecture au laboratoire ou copiez-la en Markdown en un clic.

Tester sur mon manuscrit ↑
YILA RESEARCH WORKBENCH · FIELD REFERENCE

Eleven scholarly AI-writing tells to audit before submission, paired with what must stay untouched.

Protocole 6 couches
2.1Over-claiming verbsprove, demonstrate, guarantee; bare "significantly"Use shows / improves by 4–7 pts (paired test p < 0.01, Table 3).
2.2Significance hypepaves the way for, paramount importance, sheds light onName the exact failure mode or bottleneck solved.
2.3Empty intensifiersextensive / comprehensive experiments, a wide range ofState the exact count and named benchmarks (e.g., three datasets).
2.4Novelty paddingrepeated "novel", "to the best of our knowledge"State precisely what prior offline or single-site work omitted.
2.5Formulaic openers"In recent years, X has attracted increasing attention..."Open directly with the domain constraint or structural problem.
2.6Connective stackingMoreover / Furthermore / Additionally on consecutive sentencesDrop mechanical transitions; let proposition order carry logic.
2.7Contribution clichés(1) a novel method; (2) extensive experiments; (3) strong resultsName the mechanism, the quantified robustness margin, and the artifact.
2.8Citation dumping"Many methods exist [3, 7, 9, 12, 15]."Anchor the closest prior work and state how yours differs; keep all cite keys.
2.9Hedging by vaguenesssomewhat, relatively, fairly, to some extentQuantify variance across seeds/subgroups or cut the empty adverb.
2.10Boilerplate emphasis"It is worth noting that", "Notably,", "Importantly,"Delete the throat-clearing; let the finding carry its own weight.
2.11Clause-stacked sentences & em-dashes35+ word chains of which/that/while + AI em-dashes (—)Split into crisp single-claim sentences; recast em-dashes cleanly.

Conventions scientifiques à PRÉSERVER absolument

  • Calibrated hedging ("suggests", "is consistent with", "may indicate") when causality is uncertain
  • Passive voice in Methods & protocols ("Samples were normalized to total protein")
  • Standard first-person plural ("we"), formal definitions, symbols, equations, and every cite key
Yila Research Workbench · yila.ai/playbooks/academic-humanizeryila.ai

Six couches de rigueur éditoriale scientifique

Exécuté uniquement lorsque vous sélectionnez explicitement ce Playbook.

Stage 01
General AI-tell removal
Cuts delve, underscore, tapestry, pivotal, participial -ing tails, negative parallelisms, synonym cycling, and AI em-dashes.
Stage 02
11 academic-specific tells
Targets textbook openers, novelty padding, empty intensifiers, connective stacking, citation dumping, and 30+ word clause chains.
Stage 03
Scholarly convention shield
Preserves calibrated hedging ("is consistent with"), passive voice in Methods, "we", formal notation, and 100% of numbers and citations.
Stage 04
Claim ↔ evidence matching
Ensures no verb outruns its supporting table, figure, p-value, or attributed range—flagging gaps instead of inventing data.
Stage 05
Personal voice & venue register
Reads your attached prior papers to match sentence cadence and adapts register to ICLR/NeurIPS, Nature/PNAS, or IEEE.
Stage 06
Funding-proposal mode
Tailored for NIH Specific Aims, NSF Project Summaries, and NSFC: keeps bold vision, enforces parallel outcome-aims, and ties promises to feasibility.

Intégrité scientifique et déclaration d’usage de l’IA

Ce flux améliore la clarté, l’adéquation preuve-affirmation et la voix de l’auteur sans jamais inventer de résultats ni de références. Respectez les règles de déclaration de votre revue ou agence.

Pourquoi la rédaction scientifique exige plus qu’un reformulateur générique

Conçu pour les manuscrits évalués par les pairs et les appels à projets exigeants.

CritèreReformulateurs / humaniseurs génériquesPrompt isolé dans un chat (« Enlève le style IA »)Yila Humaniseur académique
Numbers, equations & cite keysFrequently mutates technical terms, rounds numbers, or breaks [cite] keysOften drops secondary clauses or silently alters p-values on long passagesStrict invariance shield + explicit 0-drift verification in the audit report
Claim ↔ evidence calibrationSwaps synonyms blindly without checking tables, comparators, or statisticsEither leaves over-claims ("prove") intact or over-hedges clear resultsMatches verb strength to exact ranges & flags unbacked claims for the author
Personal author voice matchingForces a casual, blog-like register unfit for journalsDefaults to generic assistant prose across turnsUpload 1–2 of your prior papers to mirror your rhythm, transitions & hedging
Grant proposal mode (NSF / NIH / NSFC)Flattens the forward-looking vision reviewers scoreApplies paper-style trimming that weakens Specific Aims & SummariesDedicated Grant Mode keeps vision, parallelizes aims & anchors feasibility
Full-document deliverablesPlain text paste box onlyTruncates long manuscripts in chatAccepts .docx, .tex, .pdf & .md; exports clean files + structured audit log

Questions fréquentes

Will Academic Humanizer ever change my experimental numbers, p-values, equations, or citations?

No. Every number, unit, statistical value, equation, symbol, formal definition, and citation key stays untouched. When a strong verb like "prove" or "significantly" lacks quantitative backing in your draft, Yila calibrates the verb or flags the sentence in the audit log so you can insert your table or test statistic—never inventing data.

Will Yila trigger this Playbook automatically during my normal research chats?

No. Academic Humanizer is strictly explicit-only. It runs if and only if you explicitly select it from the top navigation menu, the composer chip bar, or this page.

How does Personal Voice Matching work?

Click "Attach your prior paper (match voice)" in the composer and upload one or two published PDFs, Word files, or LaTeX files you wrote yourself. Yila analyzes your sentence-length distribution, active/passive balance, transition economy, and hedging habits before rewriting your new draft.

Why does Grant Proposal Mode treat proposals differently from papers?

A paper reports completed evidence, so sweeping ambition is trimmed. A grant proposal (NIH Specific Aims, NSF CAREER, NSFC) is scored on vision plus feasibility: trimming its ambition weakens the pitch. Grant Mode preserves long-term vision, concentrates editing on the decisive opening pages, converts method-as-aim bullets into parallel outcome-framed aims, and ties bold claims to preliminary data or team track record.

How is this different from the quick AI Humanizer under Tools?

The standalone tool under Tools performs fast general-purpose paraphrasing per 100 words. This Playbook runs Yila’s full 6-layer scholarly editing workflow inside the Research Workbench—supporting .docx/.tex/.pdf manuscripts, personal voice-reference papers, Grant Proposal Mode, and a transparent change audit.

Can I continue revising or verifying citations in the same workbench task afterward?

Yes. Because the run lives inside your Yila Research Workbench, you can immediately follow up to adjust register for another journal, verify flagged citations against OpenAlex/PubMed/Crossref, or export updated Word and LaTeX files.

Prêt à redonner à votre brouillon votre véritable voix scientifique ?

Collez un passage ou joignez votre manuscrit ci-dessus pour lancer la passe en 6 couches.