TL;DR

In our new paper, we demonstrate that we can achieve selective generalisation of misalignment by midtraining Nemotron 120B on synthetic documents describing how AIs can be misaligned in a special <quarantine_token> mode, indicated by a new special token (a neologism), but are otherwise aligned outside this mode. We find positive results for SFT and on-policy RL post-training. However, the technique is sensitive: it is sensitive to training hyperparameters, suffers from conditional misalignment, exhibits perplexing scaling trends, and mostly underperforms vanilla Inoculation Prompting. While not a production-ready intervention, we view this as the groundwork for future interventions that enable us to guide post-training-induced misalignment via base model data curation.

This post provides a high-level summary. We abstract away many details and exclude numerous experiments. We encourage readers to read our paper for more details.

Authors: Kyle O'Brien¹, Edward James Young¹, Puria Radmard¹, Nathalie Kirch¹, Cameron Tice¹, Tomek Korbak², David Demitri Africa³

¹Geodesic Research — ²OpenAI — ³UK AI Security Institute

This work was conducted by Geodesic Research and advisors from OpenAI and UK AISI

Abstract: Large language models (LLMs) often learn both desirable and undesirable properties during post-training. We study whether midtraining, an earlier training stage [...]

---

Outline:

(00:13) TL;DR

(03:06) Method

(05:41) Results

(10:00) Discussion

(12:41) Acknowledgements

(12:45) Community: This work was improved through discussions with many members of the community. Any omissions are the unintentional fault of the authors alone. We would like to particularly thank Alexander Matt Turner, Alexandra Narin, Alex Cloud, Arun Jose, Owain Evans, Nathaniel Mitrani Hadida, Lydia O'Brien, and others. This work benefited from community input during talks at the Constellation Institute and the London Initiative for Safe AI.

(13:15) Resources: This work was made possible only by the generous support of the UK AI Security Institute in granting access to the Isambard AI Compute Cluster. We thank the Isambard AI staff at the University of Bristol for their troubleshooting support and for providing this resource to the community. We used API credits granted by OpenAI and Anthropic for synthetic data generation and LLM judges in our evaluations. Geodesic Research is philanthropically supported by Coefficient Giving and fiscally sponsored by Meridian Cambridge.

The original text contained 2 footnotes which were omitted from this narration.

---

First published:
September 15th, 2026

Source:
https://www.lesswrong.com/posts/o4Jmyn25TWm8jRAy8/inoculation-midtraining-with-learned-neologisms

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Figure 1: The problem of selective generalisation.

Figure 2: Our Approach to Inoculation Midtraining.

Figure 3: Representative Inoculation Midtraining Document.

Figure 4: Inoculation Midtraining reduces average misalignment induced by SFT on risky advice.

Figure 5: Inoculation Midtraining generalises benign data properties.

Figure 7: Inoculation Midtraining confines misalignment learned from our simpler RL environment, on par with inoculation prompting.

Figure 12: Inoculation Midtraining exhibits non-monotonic scaling with increased training data.

Figure 14: Inoculation Midtraining is Sensitive to Prompting.

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

“Inoculation Midtraining with Learned Neologisms” by Kyle O’Brien, Edward James Young, Puria, Nathalie Kirch, Cam, Tomek Korbak, David Africa

LessWrong (30+ Karma)

“Inoculation Midtraining with Learned Neologisms” by Kyle O’Brien, Edward James Young, Puria, Nathalie Kirch, Cam, Tomek Korbak, David Africa