# google-deepmind/detectorscope

> A toolkit for building AI-text detectors whose decisions are mechanistically grounded and auditable: each prediction is a sum of named features localized in the text. Built on sparse autoencoders and a concept bottleneck over the language model's residual stream.

- Magnitude: 5.4 out of 10 — Early signal
- Stars: 143 total · +24 stars measured 2026-10-28, 07:04–09:13 UTC, ≈ 73 by evening
- Star trust: star growth looks organic
- Category: Language models · Language: Python · License: Apache-2.0 · Created: 2026-10-24 · Last push: 2026-10-24
- GitHub: https://github.com/google-deepmind/detectorscope · Homepage: https://aclanthology.org/2026.emnlp-main.762/ · Page: https://gitnova.dev/en/r/google-deepmind/detectorscope

## Useful for

- Check a text for AI authorship and see which phrases drove the decision
- Convert your own black-box detector into an auditable one via the pipeline
- Compare tendencies of different LLMs by analyzing their outputs through detector features

## Why it’s here

- Star-counter measurements on 2026-10-28 (UTC), 07:04–09:13: 119 → 143 stars (+24). This is the change over that interval.
- Estimated end-of-day forecast: about +73 stars, using observed gains and the previous day.
- The repository is 4 days old and already has 143 stars. With less than two weeks of history, there's no usual pace to compare the spike against yet.
- Top new repositories this week: #135.
- Hacker News: “DetectorScope: AI text detectors that show their work” — 9 points, 3 days ago.
- Recent forks include notable developers: @bigsnarfdude (322 followers).

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 13
- Issues and pull requests: 1
- Watchers: 0
- Average over the last week: 38 per day
- Usual pace: too little history (under two weeks)
- Stars in the last hour (measured): 10

## Stars per day, last 5 days (oldest → newest, today is partial)

2026-10-24 … 2026-10-28: 8, 2, 15, 93, 24

## Hacker News

- DetectorScope: AI text detectors that show their work — 9 points, 2 comments: https://news.ycombinator.com/item?id=50222061

## Spotted in now

- Top new repositories this week: #135

## More in this category

1. **togg53192-cmd/jailbreaks** — 7.4 · Breakout · Language models · +671 stars measured 2026-10-28, 00:02–09:12 UTC, ≈ 1,561 by evening
   A collection of 31 system prompts, persona locks, and AGENTS.md files for 25+ models (Claude, Gemini, Grok, GLM, DeepSeek, Kimi, Qwen, Gemma, GPT-OSS). Intended for research on how instruction hierarchies and policy overrides affect model…
   Full card: https://gitnova.dev/en/r/togg53192-cmd/jailbreaks.md
2. **SoMarkAI/OXR** — 7.4 · Breakout · Language models · Python · +150 stars measured 2026-10-28, 03:32–09:12 UTC, ≈ 511 by evening
   OXR is an open-source document parsing model that turns PDFs and document images into Markdown or JSON with layout coordinates, recognizing 13 element types such as tables, formulas, chemical structures and code. With 1.3B parameters it…
   Full card: https://gitnova.dev/en/r/SoMarkAI/OXR.md
3. **NVIDIA/auto-ontology** — 6.1 · Early signal · Language models · Python · +87 stars measured 2026-10-28, 00:02–09:12 UTC, ≈ 210 by evening
   NVIDIA's Python framework that builds a semantic ontology graph over enterprise metadata, giving LLM agents trusted metric and entity definitions for reliable text-to-SQL.
   Full card: https://gitnova.dev/en/r/NVIDIA/auto-ontology.md
4. **lyramakesmusic/neuron-explorer** — 5.9 · Peaking · Language models · Python · +50 stars measured 2026-10-28, 00:01–09:13 UTC, ≈ 148 by evening
   A steering-first neuron explorer for Qwen3.5-4B: it pushes every MLP neuron up and down on prompts, keeps all rollouts, and lays the results out over a 3D view of the model for interpretability research.
   Full card: https://gitnova.dev/en/r/lyramakesmusic/neuron-explorer.md
5. **baidu/Unlimited-OCR** — 5.8 · Breakout · Language models · Python · +219 stars measured 2026-10-28, 00:02–09:13 UTC, ≈ 559 by evening
   Baidu's OCR model for document parsing: extracts text from single images, multi-page files and PDFs in one pass, supporting long context up to 32768 tokens.
   Full card: https://gitnova.dev/en/r/baidu/Unlimited-OCR.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-28 09:25 UTC, updated every 30 minutes.