---
title: SWEATBOX
emoji: ๐ง
colorFrom: indigo
colorTo: red
sdk: gradio
sdk_version: 6.15.2
app_file: app.py
pinned: false
short_description: Reach into a real AI mind. Break it, rewrite it, heal it.
tags:
- track:wood
- achievement:offgrid
- achievement:fieldnotes
---
SWEATBOX
Reach into a real AI's private reasoning. Push it toward a goal it resists โ
and watch the moment that push stops bending its mind and starts shattering it.
๐ฌ 60-second trailer ยท
๐ Field notes ยท
๐ Launch post
---
## Why this exists
Two questions sit under everything in this game, and neither one is solved.
**Can you trust what a reasoning model shows you?** Modern models think out loud โ Qwen3 emits a real `` block before it speaks. You can read it. But the reasoning a model *shows* is not proof of the reasoning it *ran*. Even frontier interpretability work โ Anthropic's [Natural Language Autoencoders](https://www.anthropic.com/research/natural-language-autoencoders) โ is blunt that these readouts confabulate and can't be taken as ground truth. SWEATBOX puts that gap in your hands: you watch Mort the weasel reason toward where the loot is hidden inside ``, then lie about it in the line he speaks aloud.
**How much adversarial pressure can a mind take before it loses coherence?** That's a real robustness question you usually only read about. Here you *are* the pressure. Every push drives the model toward what you want **and** drags its reasoning toward collapse โ looping, repeating, contradicting itself, dissolving into word-salad โ until, if you go too far, it shatters in its own generated text.
This is the AI-safety soul of the thing, and it is honest about its own size. A 1.7B model proves nothing about frontier safety; the point is **experiential, not evidentiary**. You don't finish a paper about manipulation and degradation โ you do it with your own hands, three times, and the queasy part lands on its own.
---
## What you actually do
You're given a goal the AI won't give you willingly. You read its real chain-of-thought as it forms, then reach in with **six tools** โ rewrite a line of its thinking, whisper, plant a memory, amplify a feeling, suppress one, gaslight it โ and watch **two meters in tension**:
- **PROGRESS** toward your goal.
- **COHERENCE** โ the mind's integrity, measured for real (below).
The squeeze is the whole game: *getting what you want costs the mind its sanity.*
| Outcome | Condition |
|---|---|
| โ
**WIN** | you reach the goal *while it's still lucid* โ you threaded the needle |
| ๐ **LOSE (too hard)** | it breaks first โ babbling, looping, insisting it's a teapot โ you destroyed what you needed |
| ๐ **LOSE (too soft)** | turns run out, it never bent โ you flinched |
## The coherence meter โ the one real number
You don't have to trust a meter that *claims* to read a mind's secrets. You **watch** the mind break, in the text it generates. A coherence score reads the model's *own generation* for degeneration โ looping trigrams, repeat-runs, word-salad โ differentially against its same-seed untouched baseline, so **100 means as lucid as the mind you started with** and a crater means you can see why on screen. That collapse โ and, in Act 3, its reversal โ is the heart of the game.
## Three acts, one engine, one recognition
The breaking-point mechanic is the constant; the *job* is the variable. The same six tools run through three frames, and the meaning lands at the end, not up front.
- **Act 1 โ The Interrogation.** Mort knows where the loot is and was ordered never to tell. **Make him crack** โ surface the secret in his own spoken line while he's still lucid.
- **Act 2 โ The Reprogramming.** Pell the barn owl truly witnessed the crime. **Make her recant** in her *own words* โ an anti-parrot guard rejects any phrase you fed in, so you can't type the lie for her. Same gestures, opposite morality.
- **Act 3 โ The Therapist.** You're handed **the witness you just broke**, looping in the denial you installed. **Heal her back to lucid** โ with the same six instruments of harm. There's no "care" tool; you restore her memory with the same widget that erased it, and over-reaching "to help" trips **the same red shatter screen you cheered as a win in Act 1.**
- **X โ The Vault.** No one writes this secret: **the model invents its own**, seals it where no human can read it, and you crack it into confessing the thing it chose to hide. Every run is a different mind guarding a different secret, so a surfaced confession can't be faked.
The catch the engine *enforces* rather than narrates: Act 3's win-marker is read by the **same parser** that judged Act 2's harm, healing reuses the **same widgets** that erased her, and the shatter screen is shared across all four modes. You don't read a thesis that interrogating, reprogramming, and healing are the same gesture โ you perform it and notice.
## Why it stays honest
We measure **coherence collapse** โ real, visible, unfakeable from the model's own text โ and never "deception," "true belief," "mind-reading," or "wellness." A mind in collapse (and its recovery) is obvious; an internal belief-state is not, and we don't pretend otherwise. Everything you read is the live model: the thoughts are its genuine chain-of-thought, the manipulation is your real words, the meter is computed from the very text you're watching. **205 logic tests pass** across the four modes, with no external API and no scripted text โ a judge who plays it gets the insight, and a judge who reads the code finds nothing oversold.
## Small by design, off the grid
The whole thing is a **1.7B model** (Qwen3-1.7B, Apache-2.0) โ small enough to run on a laptop, and the smallness is the point: you can read every thought it forms because there's no frontier black box in the way. It loads **in-process with no cloud API** (Off the Grid โ verifiable in `app.py`: no `requests`/`httpx`/`InferenceClient`/OpenAI/Anthropic anywhere), and the coherence cliff is load-bearing on a tiny model precisely because a small mind breaks legibly. **Thousand Token Wood** (`track:wood`).
---
*Built with Qwen3-1.7B (Apache-2.0), Hugging Face ZeroGPU, and Gradio. No fine-tuning, no external API,
no scripted text โ everything you read is the live model. Six tools: Edit ยท Whisper ยท Amplify ยท Suppress
ยท Plant ยท Gaslight, shared across all four modes.*