docs: integrate comprehensive AI research and feature matrix#2
docs: integrate comprehensive AI research and feature matrix#2edithatogo merged 1 commit intomainfrom
Conversation
📝 WalkthroughWalkthroughDocumentation expansion adds a "SIGNS OF AI WRITING MATRIX" section mapping AI-writing detection patterns across platforms (Wikipedia, GPTZero, Originality.ai, Copyleaks, Winston AI, Turnitin) to ten markdown files, with corresponding build script update to include the new pattern matrix file in compiled outputs. Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~10 minutes Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Summary of ChangesHello @edithatogo, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request significantly upgrades the project's documentation by introducing a comprehensive matrix that details various patterns of AI-generated text and cross-references them with major AI detection platforms and academic sources. This enhancement aims to centralize and clarify information on AI writing detection, making it more accessible and robust for users. Highlights
🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console. Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here. You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension. Footnotes
|
PR Compliance Guide 🔍Below is a summary of compliance checks for this PR:
Compliance status legend🟢 - Fully Compliant🟡 - Partial Compliant 🔴 - Not Compliant ⚪ - Requires Further Human Verification 🏷️ - Compliance label |
|||||||||||||||||||||||||
PR Code Suggestions ✨Explore these optional code suggestions:
|
|||||||||||
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces a comprehensive AI writing detection matrix, integrating it across all relevant documentation files. The use of a modular source file (pattern_matrix.md) and a build script (sync-adapters.js) to propagate the changes is an excellent approach for maintaining consistency. The new documentation is well-structured and a valuable addition. I have one suggestion for the build script to improve its performance and maintainability.
|
|
||
| let full = header.replace('<<<<[CORE_FRONTMATTER]>>>>', coreFM); | ||
| full = full + '\n' + corePatterns + '\n' + researchRefs; | ||
| full = full + '\n' + corePatterns + '\n' + researchRefs + '\n' + patternMatrix; |
There was a problem hiding this comment.
For improved readability and maintainability, you could use an array for the content parts and join them. This makes it easier to manage the list of content to be appended.
As a broader performance note, consider reading these content files once outside the compileSkill function, since it's called multiple times with the same content files (except for headerPath).
| full = full + '\n' + corePatterns + '\n' + researchRefs + '\n' + patternMatrix; | |
| full = full + '\n' + [corePatterns, researchRefs, patternMatrix].join('\n'); |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 8
🤖 Fix all issues with AI agents
In `@adapters/antigravity-skill/SKILL_PROFESSIONAL.md`:
- Around line 658-683: Add blank lines before and after the new headings to
satisfy MD022: ensure there's an empty line above and below "## SIGNS OF AI
WRITING MATRIX" and likewise around the "### Source Definitions" heading so each
heading is separated from surrounding content by a single blank line; update
SKILL_PROFESSIONAL.md by inserting those blank lines around those heading
tokens.
In `@adapters/antigravity-skill/SKILL.md`:
- Around line 663-688: The heading "## SIGNS OF AI WRITING MATRIX" in SKILL.md
is missing the required blank line after it (MD022); insert a single empty line
immediately following that heading so the section starts with a blank line
before the table content, ensuring the heading and the subsequent table/content
are separated (look for the exact heading string "## SIGNS OF AI WRITING MATRIX"
to locate where to add the blank line).
In `@adapters/copilot/COPILOT.md`:
- Around line 679-686: MD022: The "Source Definitions" heading is missing a
blank line after it which violates markdownlint; update the markdown around the
"Source Definitions" heading by inserting a single blank line immediately after
the heading line (the line containing "### Source Definitions") so there is a
separation between the heading and the following list items, ensuring the
table/paragraph above remains intact and no other lines are modified.
In `@adapters/qwen-cli/QWEN.md`:
- Around line 679-686: Markdown lint MD022: add a blank line after the heading
"Source Definitions" to satisfy the rule; locate the "### Source Definitions"
heading and insert one empty line immediately below it so there is a blank line
separating the heading from the following list items.
In `@adapters/vscode/HUMANIZER.md`:
- Around line 679-686: The markdown linter flagged MD022 for the "Source
Definitions" heading; add a blank line immediately after the heading "### Source
Definitions" to separate it from the following list; locate the heading in
HUMANIZER.md and insert one empty line between the heading and the first list
item so the heading is properly surrounded by blank lines.
In `@SKILL_PROFESSIONAL.md`:
- Around line 645-672: The Markdown headings added ("SIGNS OF AI WRITING MATRIX"
and "Source Definitions") need blank lines above and below to satisfy MD022;
update SKILL_PROFESSIONAL.md by inserting a single blank line before each
heading and one blank line after the heading lines (or before the following
content/paragraph) so the headings are separated from surrounding blocks and
linter warnings are resolved.
In `@SKILL.md`:
- Around line 650-677: Add missing blank lines around the headings to satisfy
MD022: insert a blank line before the "## SIGNS OF AI WRITING MATRIX" heading
and a blank line between that heading and the following table, and likewise
ensure there is a blank line before the "### Source Definitions" heading and a
blank line after it (i.e., separate the heading from surrounding text/table).
In `@src/pattern_matrix.md`:
- Around line 2-28: The matrix "SIGNS OF AI WRITING MATRIX" currently shows
undocumented [x] marks for tools like GPTZero, Originality.ai, Copyleaks,
Winston AI, and Turnitin; add a legend ("[x] = explicitly documented by source;
[ ] = not documented or not found") under the heading, and update the table
entries to reflect only explicitly supported signals by each vendor (change
unsupported [x] to [ ] or mark as "inferred"); then add a "Sources" section
listing citations (blog posts, technical docs, FAQs) for each vendor name
(GPTZero, Originality.ai, Copyleaks, Winston AI, Turnitin) that justify the
remaining [x] marks or note which entries are inferred.
| ## SIGNS OF AI WRITING MATRIX | ||
|
|
||
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | ||
|
|
||
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | ||
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | ||
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | ||
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | ||
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | ||
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | ||
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | ||
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | ||
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. |
There was a problem hiding this comment.
Fix missing blank lines around headings to satisfy MD022.
Markdownlint flags missing blank lines around the new headings; add an empty line before/after the headings to keep lint clean.
🧹 Suggested edit
-## SIGNS OF AI WRITING MATRIX
-
-The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them.
+## SIGNS OF AI WRITING MATRIX
+
+The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them.
@@
-### Source Definitions
-- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).
+### Source Definitions
+
+- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).🧰 Tools
🪛 markdownlint-cli2 (0.20.0)
[warning] 659-659: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
[warning] 682-682: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
🤖 Prompt for AI Agents
In `@adapters/antigravity-skill/SKILL_PROFESSIONAL.md` around lines 658 - 683, Add
blank lines before and after the new headings to satisfy MD022: ensure there's
an empty line above and below "## SIGNS OF AI WRITING MATRIX" and likewise
around the "### Source Definitions" heading so each heading is separated from
surrounding content by a single blank line; update SKILL_PROFESSIONAL.md by
inserting those blank lines around those heading tokens.
| ## SIGNS OF AI WRITING MATRIX | ||
|
|
||
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | ||
|
|
||
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | ||
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | ||
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | ||
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | ||
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | ||
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | ||
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | ||
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | ||
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. |
There was a problem hiding this comment.
Add a blank line after the new heading to clear MD022.
Markdownlint reports missing blank lines around the heading; add an empty line after the heading.
🧹 Suggested edit
-### Source Definitions
-- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).
+### Source Definitions
+
+- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| ## SIGNS OF AI WRITING MATRIX | |
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | |
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | |
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | |
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | |
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | |
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | |
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | |
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | |
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| ### Source Definitions | |
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | |
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | |
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | |
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | |
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | |
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. | |
| ## SIGNS OF AI WRITING MATRIX | |
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | |
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | |
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | |
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | |
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | |
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | |
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | |
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | |
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| ### Source Definitions | |
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | |
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | |
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | |
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | |
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | |
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. |
🧰 Tools
🪛 markdownlint-cli2 (0.20.0)
[warning] 682-682: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
🤖 Prompt for AI Agents
In `@adapters/antigravity-skill/SKILL.md` around lines 663 - 688, The heading "##
SIGNS OF AI WRITING MATRIX" in SKILL.md is missing the required blank line after
it (MD022); insert a single empty line immediately following that heading so the
section starts with a blank line before the table content, ensuring the heading
and the subsequent table/content are separated (look for the exact heading
string "## SIGNS OF AI WRITING MATRIX" to locate where to add the blank line).
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". |
There was a problem hiding this comment.
Fix MD022: add blank line around “Source Definitions.”
markdownlint flags missing blank lines around the heading (Line 682). Add a blank line after the heading.
✏️ Suggested fix
-### Source Definitions
-- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).
+### Source Definitions
+
+- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | |
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| ### Source Definitions | |
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | |
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | |
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | |
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | |
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | |
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| ### Source Definitions | |
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | |
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | |
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | |
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". |
🧰 Tools
🪛 markdownlint-cli2 (0.20.0)
[warning] 682-682: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
🤖 Prompt for AI Agents
In `@adapters/copilot/COPILOT.md` around lines 679 - 686, MD022: The "Source
Definitions" heading is missing a blank line after it which violates
markdownlint; update the markdown around the "Source Definitions" heading by
inserting a single blank line immediately after the heading line (the line
containing "### Source Definitions") so there is a separation between the
heading and the following list items, ensuring the table/paragraph above remains
intact and no other lines are modified.
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". |
There was a problem hiding this comment.
Fix MD022: add blank line around “Source Definitions.”
markdownlint flags missing blank lines around the heading (Line 682). Add a blank line after the heading.
✏️ Suggested fix
-### Source Definitions
-- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).
+### Source Definitions
+
+- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).🧰 Tools
🪛 markdownlint-cli2 (0.20.0)
[warning] 682-682: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
🤖 Prompt for AI Agents
In `@adapters/qwen-cli/QWEN.md` around lines 679 - 686, Markdown lint MD022: add a
blank line after the heading "Source Definitions" to satisfy the rule; locate
the "### Source Definitions" heading and insert one empty line immediately below
it so there is a blank line separating the heading from the following list
items.
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". |
There was a problem hiding this comment.
Fix MD022: add blank line around “Source Definitions.”
markdownlint flags missing blank lines around the heading (Line 682). Add a blank line after the heading.
✏️ Suggested fix
-### Source Definitions
-- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).
+### Source Definitions
+
+- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).🧰 Tools
🪛 markdownlint-cli2 (0.20.0)
[warning] 682-682: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
🤖 Prompt for AI Agents
In `@adapters/vscode/HUMANIZER.md` around lines 679 - 686, The markdown linter
flagged MD022 for the "Source Definitions" heading; add a blank line immediately
after the heading "### Source Definitions" to separate it from the following
list; locate the heading in HUMANIZER.md and insert one empty line between the
heading and the first list item so the heading is properly surrounded by blank
lines.
|
|
||
|
|
||
| ## SIGNS OF AI WRITING MATRIX | ||
|
|
||
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | ||
|
|
||
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | ||
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | ||
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | ||
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | ||
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | ||
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | ||
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | ||
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | ||
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. |
There was a problem hiding this comment.
Add blank lines around new headings to clear MD022.
Markdownlint flags the heading spacing; add blank lines around the new headings to match the style rules.
🧹 Suggested edit
-## SIGNS OF AI WRITING MATRIX
-
-The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them.
+## SIGNS OF AI WRITING MATRIX
+
+The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them.
@@
-### Source Definitions
-- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).
+### Source Definitions
+
+- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).🤖 Prompt for AI Agents
In `@SKILL_PROFESSIONAL.md` around lines 645 - 672, The Markdown headings added
("SIGNS OF AI WRITING MATRIX" and "Source Definitions") need blank lines above
and below to satisfy MD022; update SKILL_PROFESSIONAL.md by inserting a single
blank line before each heading and one blank line after the heading lines (or
before the following content/paragraph) so the headings are separated from
surrounding blocks and linter warnings are resolved.
|
|
||
|
|
||
| ## SIGNS OF AI WRITING MATRIX | ||
|
|
||
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | ||
|
|
||
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | ||
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | ||
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | ||
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | ||
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | ||
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | ||
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | ||
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | ||
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. |
There was a problem hiding this comment.
Fix MD022 around matrix headings.
markdownlint reports missing blank lines around “SIGNS OF AI WRITING MATRIX” and “Source Definitions” (Line 650, Line 654, Line 677). Add blank lines before/after these headings.
✏️ Suggested fix
-## SIGNS OF AI WRITING MATRIX
-
-The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them.
+## SIGNS OF AI WRITING MATRIX
+
+The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them.
-### Source Definitions
-- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).
+### Source Definitions
+
+- **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup).📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| ## SIGNS OF AI WRITING MATRIX | |
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | |
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | |
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | |
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | |
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | |
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | |
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | |
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | |
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| ### Source Definitions | |
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | |
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | |
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | |
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | |
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | |
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. | |
| ## SIGNS OF AI WRITING MATRIX | |
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | |
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | |
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | |
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | |
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | |
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | |
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | |
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | |
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | |
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | |
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | |
| ### Source Definitions | |
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | |
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | |
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | |
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | |
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | |
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. |
🧰 Tools
🪛 markdownlint-cli2 (0.20.0)
[warning] 650-650: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
[warning] 654-654: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
[warning] 677-677: Headings should be surrounded by blank lines
Expected: 1; Actual: 0; Below
(MD022, blanks-around-headings)
🤖 Prompt for AI Agents
In `@SKILL.md` around lines 650 - 677, Add missing blank lines around the headings
to satisfy MD022: insert a blank line before the "## SIGNS OF AI WRITING MATRIX"
heading and a blank line between that heading and the following table, and
likewise ensure there is a blank line before the "### Source Definitions"
heading and a blank line after it (i.e., separate the heading from surrounding
text/table).
| ## SIGNS OF AI WRITING MATRIX | ||
|
|
||
| The following matrix maps observed patterns of AI-generated text to the major detection platforms and academic resources that document them. | ||
|
|
||
| | Pattern Category | Specific Signs | Wikipedia | GPTZero | Originality.ai | Copyleaks | Winston AI | Turnitin | | ||
| | :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | | ||
| | **Statistical** | **Low Perplexity** (Predictable word choices) | [x] | [x] | [x] | [x] | [x] | [x] | | ||
| | | **Uniform Burstiness** (Consistent rhythms) | [x] | [x] | [ ] | [x] | [x] | [ ] | | ||
| | **Stylistic** | **Repetitive Phrasing** / Sentence starts | [x] | [x] | [x] | [x] | [x] | [ ] | | ||
| | | **Lack of Emotion / Nuance / Voice** | [x] | [ ] | [x] | [x] | [x] | [ ] | | ||
| | | **Copula Avoidance** ("serves as" vs "is") | [x] | [ ] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Over-Significance / Inflation** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | **Grammar** | **Flawless / Hyper-Correct Grammar** | [x] | [x] | [ ] | [ ] | [ ] | [ ] | | ||
| | | **Tautology / Redundant Stating** | [x] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Technical** | **Factual "Fumbles" / Hallucinations** | [x] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
| | | **Unicode / Hidden Text Artifacts** | [ ] | [ ] | [x] | [ ] | [ ] | [ ] | | ||
| | **Advanced** | **Bypasser / Paraphraser Detection** | [ ] | [x] | [ ] | [ ] | [ ] | [x] | | ||
| | | **Semantic Conceptual Patterns** | [ ] | [ ] | [ ] | [x] | [ ] | [ ] | | ||
|
|
||
| ### Source Definitions | ||
|
|
||
| - **Wikipedia:** Community-maintained "Signs of AI writing" (WikiProject AI Cleanup). | ||
| - **GPTZero:** Focuses on statistical randomness (Perplexity) and variation (Burstiness). | ||
| - **Originality.ai:** Targets content marketing spam, tautology, and technical evasion. | ||
| - **Copyleaks:** Emphasizes semantic conceptual analysis and "Stylistic Markers". | ||
| - **Winston AI:** Scans for structural rhythm inconsistencies and predictable patterns. | ||
| - **Turnitin:** Focuses on prose likelihood and detection of "AI Bypasser" tool signatures. |
There was a problem hiding this comment.
🧩 Analysis chain
🌐 Web query:
GPTZero official documentation detection methods perplexity burstiness
💡 Result:
GPTZero’s official docs describe a multi-component AI-detection system where perplexity and burstiness are part of (and historically the “first layer” of) its approach, alongside additional techniques like deep learning sentence classification (and other components in a “7-component / seven-layer” model). (gptzero.me)
Perplexity (in GPTZero)
- GPTZero explains perplexity as a measure of how predictable the text is to a language model—i.e., how likely an AI model would have produced the same wording. Lower perplexity is presented as more AI-like; higher perplexity as more human-like. (gptzero.me)
Burstiness (in GPTZero)
- GPTZero describes burstiness as measuring variation across the document (e.g., variation in sentence construction/complexity over time), contrasting humans’ tendency to vary more with models’ more uniform, “formulaic” patterns—so lower burstiness is presented as more AI-like and higher burstiness as more human-like. (gptzero.me)
“Beyond perplexity & burstiness” (what GPTZero says it also uses)
- GPTZero states it has evolved into a multilayer system with seven components, not relying only on these two statistics. (gptzero.me)
- Its technology pages describe deep-learning, end-to-end approaches and a sentence-by-sentence classifier that outputs probability/confidence at sentence level (and it can report at sentence/paragraph/document levels). (gptzero.me)
Citations:
- 1: https://gptzero.me/news/perplexity-and-burstiness-what-is-it/?utm_source=openai
- 2: https://gptzero.me/news/perplexity-and-burstiness-what-is-it/?utm_source=openai
- 3: https://gptzero.me/news/perplexity-and-burstiness-what-is-it/?utm_source=openai
- 4: https://gptzero.me/news/how-ai-detectors-work/?utm_source=openai
- 5: https://gptzero.me/technology?utm_source=openai
🌐 Web query:
Originality.ai how it detects AI writing features
💡 Result:
Originality.ai describes its AI-writing detection as a supervised text-classification model (not a simple “perplexity-only” checker):
- Core approach (what it’s doing): it uses a fine-tuned transformer-based discriminator/classifier trained to label text as human vs. AI by learning patterns from millions of labeled examples (“known AI” and “known human” text). It outputs a probability/confidence score and can highlight AI-likely sections at sentence/paragraph level. [1][2]
- Model/training details they disclose: the detector is based on a modified BERT-style architecture, and they describe pretraining inspired by ELECTRA (generator + discriminator), then fine-tuning on their AI-vs-human dataset. They also say their AI text training data is diversified across different models and different sampling settings (e.g., temperature, top‑k, nucleus sampling) and is frequently manually reviewed. [1]
- “AI writing features” (examples they discuss): in explaining common detection signals, they mention differences that detectors may exploit such as burstiness, perplexity/predictability, frequency/word-choice distributions, readability/fluency patterns, and punctuation patterns. (They present these as common feature categories used in the space, alongside the fine-tuned-model approach.) [2]
Sources:
[1] Originality.ai — “How Does AI Content Detection Work?” (Oct 18, 2025) (originality.ai)
[2] Originality.ai — “We Have 99% Accuracy in Detecting AI: Originality.ai Study” / “AI Accuracy” (originality.ai)
Citations:
- 1: https://originality.ai/blog/how-does-ai-content-detection-work
- 2: https://originality.ai/blog/ai-content-detection-accuracy
🌐 Web query:
Copyleaks AI detection methods semantic analysis
💡 Result:
Copyleaks’ AI text detection (including its “AI Logic” explainability layer) describes a multi-signal NLP/ML approach where semantic analysis is one of several components used to judge whether text is AI-influenced.
Where “semantic analysis” fits in Copyleaks’ approach
- Copyleaks explains that text detectors use NLP to examine semantics (word meanings and relationships), along with syntax/structure and context cues, to differentiate human vs. AI writing. This is the part most directly aligned with “semantic analysis.” [1]
- Their detector is designed to work on mixed content (human + AI) and can report which parts are likely AI-generated, implying analysis at the segment/span level rather than only a single document-wide label. [4]
Signals Copyleaks highlights (esp. in AI Logic) that relate to semantics
Copyleaks’ “AI Logic” feature is positioned as an explanation layer that surfaces which detectable patterns drove the decision: [3]
- AI Phrases: flags phrases that appear more frequently in AI-written text than in human writing, which is effectively phrase-level distributional/usage-pattern analysis (often correlated with semantic/lexical choice patterns). [2]
- AI Source Match: checks whether parts of the text match AI-generated content already published elsewhere, adding a provenance/reuse signal beyond purely linguistic semantics. [2]
- Patterns object with statistics + textual analysis: in the API, AI Logic can return a structured breakdown including statistical properties compared against known AI/human patterns and specific text segments exhibiting AI-like characteristics (with location/length). [3]
How Copyleaks validates the detector (method, not semantics-specific)
Copyleaks documents an internal evaluation process using labeled human vs. AI datasets, run through their API, and scored with metrics like accuracy, ROC-AUC, F1, and confusion matrices—useful context for how they claim reliability is measured. [5]
🌐 Web query:
Winston AI detection features AI writing patterns
💡 Result:
Winston AI’s AI-writing detection is built around a few visible “pattern” outputs in its reports:
- Human vs. AI probability score (%) for the whole document (i.e., likelihood the text was AI-generated vs human-written). (help.gowinston.ai)
- Sentence-by-sentence “AI prediction map” (color-coded highlights that mark individual sentences as likely AI / possible AI / likely human), so you can see where the writing looks machine-like. (cybernews.com)
- Readability scoring (text complexity), which is commonly paired with detection results to help interpret “too consistent / too polished” writing styles. (winston-ai.tawk.help)
- Writing feedback focused on structure and style (grammar, syntax, sentence structure, tone, style)—useful for identifying and revising passages that read like typical AI output. (gowinston.ai)
- OCR + file support (scan PDFs and images like JPG/PNG/TIFF to extract text and then run the same AI-pattern analysis). (quetext.com)
- Plagiarism checking alongside AI detection (separate, but often used together when reviewing authenticity). (cybernews.com)
Citations:
- 1: https://help.gowinston.ai/understanding-winston-ai/how-do-we-interpret-the-results-from-an-ai-text-scan?utm_source=openai
- 2: https://cybernews.com/ai-tools/winston-ai-review/?utm_source=openai
- 3: https://winston-ai.tawk.help/article/how-to-use-winston-ai-for-text-analysis?utm_source=openai
- 4: https://gowinston.ai/writing-feedback/?utm_source=openai
- 5: https://www.quetext.com/blog/winston-ai-detector-review-is-it-really-accurate?utm_source=openai
- 6: https://cybernews.com/ai-tools/winston-ai-review/?utm_source=openai
🌐 Web query:
Turnitin AI detection methods documentation
💡 Result:
Official Turnitin documentation (AI writing detection)
- Turnitin’s AI writing detection capabilities FAQs (English model) — explains how it works (segmentation into overlapping ~5–10 sentence chunks, per‑sentence scoring 0–1, document‑level aggregation), what it detects, training approach, supported languages, interpretation, false positives, and related features (AI paraphrasing / AI bypasser detection). Updated Nov 21, 2025. [1]
- AI writing detection model (release notes) — change log for the detection model (e.g., *% for 1–19% after July 16, 2024; Spanish/Japanese support releases; bypasser-tool detection; model updates and resubmission behavior). Updated Oct 14, 2025. [2]
- File requirements for an AI writing report — technical constraints to generate an AI report (e.g., <100MB, ≥300 words of long‑form prose, ≤30,000 words, supported languages, accepted file types). [3]
- Turnitin research/blog note on bias & minimum length — describes evaluation results and why Turnitin enforces the 300‑word minimum to keep false positives low (targeting <1%). [4]
What Turnitin says about its detection method (high level)
- It breaks a submission into overlapping segments (roughly a few hundred words / about 5–10 sentences) to score sentences in context, then assigns each sentence a 0–1 score and aggregates those scores into an overall AI percentage for “qualifying” long‑form prose. [1]
- The FAQ describes the model as using classifiers trained to detect differences in word-probability patterns between LLM-generated text and human writing, and notes the training data includes AI text plus authentic academic writing across geographies/subjects, with attention to under‑represented groups. [1]
- The release notes document ongoing expansions (e.g., Spanish, Japanese, and AI bypasser tool detection) and changes to reporting behavior (notably the *1–19% shown as % approach to reduce risk from false positives). [2]
Sources: Turnitin Guides/Help Center pages linked above. [1][2][3][4]
Add a legend and citations; most matrix claims lack explicit documentation from the platforms.
The matrix makes specific coverage claims for each detector without citations or a legend. Beyond that, official documentation from these platforms does not explicitly enumerate most of the patterns listed:
- GPTZero officially documents only perplexity and burstiness; the other [x] marks (repetitive phrasing, lack of emotion, flawless grammar) are not explicitly cited in their technical docs.
- Originality.ai mentions common detection features (burstiness, perplexity, word-choice, readability, punctuation) but does not explicitly claim tautology or emotion detection.
- Copyleaks confirms semantic analysis but does not enumerate the other patterns as distinct signals in official sources.
- Winston AI documents readability scoring and style analysis, but official docs do not describe specific detection of perplexity or the other [x] patterns listed.
- Turnitin explicitly covers bypasser detection and word-probability patterns; most other claims lack direct documentation.
Recommendations:
- Add a legend: [x] = explicitly documented by source; [ ] = not documented or not found.
- Cite official docs (blog posts, technical guides, FAQs) for each platform next to the table or in a sources section.
- Either remove unsupported [x] marks or change them to [ ] and note that the matrix reflects inferred capabilities rather than explicit platform claims.
🤖 Prompt for AI Agents
In `@src/pattern_matrix.md` around lines 2 - 28, The matrix "SIGNS OF AI WRITING
MATRIX" currently shows undocumented [x] marks for tools like GPTZero,
Originality.ai, Copyleaks, Winston AI, and Turnitin; add a legend ("[x] =
explicitly documented by source; [ ] = not documented or not found") under the
heading, and update the table entries to reflect only explicitly supported
signals by each vendor (change unsupported [x] to [ ] or mark as "inferred");
then add a "Sources" section listing citations (blog posts, technical docs,
FAQs) for each vendor name (GPTZero, Originality.ai, Copyleaks, Winston AI,
Turnitin) that justify the remaining [x] marks or note which entries are
inferred.
1 participant
User description
Integrates 10+ sources, new CSV matrix, and academic refs. Fixes blader#24, blader#27.
PR Type
Enhancement, Documentation
Description
Add AI writing detection matrix mapping patterns to detection platforms
Create modular
pattern_matrix.mdsource file for reusabilityUpdate build script to include pattern matrix in compiled outputs
Distribute matrix across all skill and adapter documentation files
Diagram Walkthrough
File Walkthrough
1 files
Add pattern matrix file to compilation pipeline11 files
Create new modular AI detection pattern matrixAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionAppend AI writing detection matrix sectionSummary by CodeRabbit
✏️ Tip: You can customize this high-level summary in your review settings.