ai gen
Research notebook / Text autoencoders
A research archive and a working plan

Reading a
sentence vector.

What survives when a whole sentence becomes one vector? Research on text autoencoders, their latent spaces, and what we can read, edit, or audit inside them.

Text autoencodersInterpretabilitySONAR
NOW / THE ACTIVE PAPER9-page target · v0.1

Understanding neuralese with TAE-Bench

The new paper project: a general testbed, a shared method comparison, and two case studies. Existing evidence, revisions, and unrun experiments are explicitly marked.

Read the new paper and experiment plan
NEXT / THE PLAN12 tasks + kanban

What should we do next?

The expanded research plan: task protocols, controls, dependencies, reading priorities and the completed multilingual pilot.

Open the plan and current kanban
01 / THE UPDATEThrough 7 Sep 2026

What’s new in text autoencoders?

The research roundup since NickyP’s February 2025 review. New models, SONAR interpretability, latent generation, and embedding inversion, with sources and scope notes.

Explore the research update
ARCHIVE / CAMPAIGN100Earlier manuscript · retained

Decodability is not abstraction

The original Campaign100 manuscript remains available. The active writing project is now the TAE-Bench paper above; this archive retains its existing results and corrections.

Read the paper
03 / THE LITERATURE20 research strands

The wider literature review

A deeper map of text-autoencoder interpretability: geometry, binding, probing, dictionaries, latent reasoning, decoder faithfulness, and the surrounding literature.

Browse the literature review
04 / THE WORKCampaign100

Done, open, and next

The experiment kanban: program tasks, claims, evidence, and follow-ups. See what has run, what has been hardened, and what still stands between the results and a paper.

Open the kanban board

Two ways into the reading

The research update asks what appeared after the February 2025 post. The literature review organizes the wider field around this project’s interpretability questions.

A working research notebook

The paper and board retain their own dated status notes. This overview connects the documents; it does not promote draft claims or mark unfinished experiments complete.

Earlier write-up: Anatomy of a Text Autoencoder.