Skip to content
AI News HubLIVE
Original source2 min read

Train on prefill, monitor at decode: Using sparse autoencoders for topic monitoring

Summary

arXiv paper introduces TopicScope, a framework for training causal topic classifiers on SAE features that assign a topic to every token during generation, generalizing from prefill to decode on Gemma-3 1B and 4B with strong robustness.

SourcearXiv AIAuthor: Rishikesh Mallagundla, Pranav Patel
Train on prefill, monitor at decode: Using sparse autoencoders for topic monitoring
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 6 Nov 2026]

Title:Train on prefill, monitor at decode: Using sparse autoencoders for topic monitoring

View a PDF of the paper titled Train on prefill, monitor at decode: Using sparse autoencoders for topic monitoring, by Rishikesh Mallagundla and 1 other authors

View PDF HTML (experimental)

Abstract:Topic monitoring is the task of labelling the topic of each token in a model's response during generation. Activation-based monitors that run at generation time have so far been trained to detect harmful content and are limited to two or three classes. We introduce TopicScope, a framework for training causal topic classifiers on SAE features that assign a topic to every token, so that the label can change partway through a response. We train three detector architectures on synthetic multi-topic conversations covering seven topics. Across Gemma-3 1B and 4B, monitors trained on features extracted from prefill hidden states were able to generalise to decoding. We also found that the classifiers are robust to perturbations, losing at most 5 macro-F1 points when off-topic asides are inserted, topics are removed or sections are reordered. Ablations conducted to understand the contribution of the SAE revealed that classifiers reaching baseline performance without it were several times larger. Our findings show that compact topic monitors can be trained inexpensively in prefill and deployed during generation. More broadly, our work shows where SAEs are useful beyond safety classification. Our code, dataset and checkpoints are available at this https URL and this https URL

Comments: 22 pages

Subjects:

Artificial Intelligence (cs.AI); Machine Learning (cs.LG)

Cite as: arXiv:2611.12935 [cs.AI]

(or arXiv:2611.12935v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2611.12935

arXiv-issued DOI via DataCite

Submission history

From: Rishikesh Mallagundla [view email] [v1] Fri, 6 Nov 2026 11:33:35 UTC (678 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Train on prefill, monitor at decode: Using sparse autoencoders for topic monitoring, by Rishikesh Mallagundla and 1 other authors

View PDF

HTML (experimental)

TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.AI

new | recent | 2026-11

Change to browse by:

cs cs.LG

References & Citations

SciX

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • TopicScope labels each token's topic during generation, allowing labels to change mid-response.
  • Monitors trained on prefill hidden-state features generalize to decoding on Gemma-3 1B and 4B.
  • Classifiers are robust to perturbations, losing at most 5 macro-F1 points under off-topic asides, topic removal, or reordering.
  • Ablations show that without SAE features, classifiers need to be several times larger to reach baseline performance.

Highlights and analysis are generated automatically and may contain errors. Check the original source.