←
AlignmentJournal38 posts
AlignmentJournal avatar

AlignmentJournal

@AlignmentJrnl

A journal for foundational AI alignment research. Submit now. Follow to see new submissions and nominate yourself to review them.

◷ Joined March 2026▧ 7 Photos and videos
AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Oct 22

The Alignment Journal publishes research on the scientific foundations of understanding, predicting, and steering artificial intelligent systems. As of today, submissions are open to everyone. What we do differently: - We pay reviewers, with bonuses for speed and insight. - Review discussion is confidential, but every accepted paper is published with a “review summary” to distill the insights generated. - Each paper is judged against our acceptance bar, not against other submissions. - Papers under review are listed publicly, and researchers can nominate themselves to review, so each paper can reach the reviewers who know the topic best. We seek work that builds theoretical or otherwise principled analyses of agency, robustness, incentives, interpretability, generalization, and long-term behavior, and whose significance survives the deletion of any particular model, dataset or benchmark. Empirical investigations of such ideas are welcome, and negative results are encouraged. We are especially interested in approaches that could apply to future superintelligent systems that interact and self-improve. You can submit your preprint in any format. Submissions are rolling and open now on the website. If you are instead interested in reviewing or serving as an action editor (basically: an area chair), create an account on our website. Links to submit, our scope and the reasoning behind the Journal are in the reply ⬇️

Post attachment
AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Nov 16

Paper: alignmentjournal.org/under-r… By making an account on our website and listing your expertise, our editors will be able to reach out to you in case a paper in your area will come in and reviewers are needed.

AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Nov 16

The Alignment Journal is looking for reviewers for "Ecology of Self-Replicating AI Agents: Collaboration Creates a Population Threshold for Takeoff". You might be interested if you have experience with population dynamics, AI cyber capability or self-replication evals, multi-agent scaling, or malware and botnet spread models. Authors stay anonymous while the paper is under review. Reviewers are paid, with a bonus for being early. Nominate yourself ⬇️

Post attachment
AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Nov 12

Now under review at the Alignment Journal: “General Agents Contain World Models, even under Partial Observability and Stochasticity”. Reviewers for this paper have been found. To review in the future, make an account on our website and list your expertise. Editors will search among the accounts when recruiting reviewers for new papers. You can also nominate yourself.

Post attachment
AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Nov 5

Now under review at the Alignment Journal: “Fragility of Value under Imperfect Alignment”. Reviewers for this paper have been found. If Goodhart’s law, proxy optimization or value fragility is your speciality, make an account on our website and list your expertise. Editors will then know to consider you when the next paper in your area comes in. ⬇️

Post attachment
Jess Riedel avatar
⟳ AlignmentJournal retweeted
Jess Riedel@Jess_Riedel·Nov 5

These are the top AI review services the Alignment Journal @AlignmentJrnl is considering incorporating. I'm very interested to hear from alignment researcher who have experience with any of them. ⋅CSPaper ⋅Google DeepReview ⋅Lemma ⋅CMU Paper Reviewer ⋅Refine

AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Oct 30

Paper: alignmentjournal.org/under-r… You can also make an account on our website to review in the future. Whenever a new paper comes in, editors will search among the accounts when recruiting reviewers.

AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Oct 30

Now under review at the Alignment Journal: "Delegating Authorization to Misaligned Agents: Coalitional Alignment and Safe Control.” Authors stay anonymous while the paper is under review. If you think this is in your area, nominate yourself to review. ⬇️

Post attachment
AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Oct 29

Paper: alignmentjournal.org/under-r… On the process: 1. Sign up on our website if you'd like to review, and list your expertise. 2. You can nominate yourself for papers in your area. 3. Editors consider nominations and search the account database when recruiting for new papers. They'll email you if there's a fit.

AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Oct 29

From today, we will be sharing the papers that come under review at the Alignment Journal, so researchers can find papers in their area they’d be interested in reviewing. First up: "Superalignment Anti-Literature Review." Authors stay anonymous while a paper is under review. ⬇️

Post attachment
AlignmentJournal avatar
AlignmentJournal@AlignmentJrnl·Oct 28

This is why the Alignment Journal scrapes everything possible from the submitted paper and auto-fills metadata on the submission form rather than collecting the information twice. Our AI checks about the paper itself (scope, anonymity, etc) are also only advisory, i.e., authors can still submit if they think the AI check got it wrong. alignmentjournal.org/submiss…

Unfortunately, submitting a paper to a journal is still a grind. Agents help, but the journal-agents now check every fiddly detail. Does the title page list keywords? Do they match the ones you typed into the portal? When checking gets cheap, u get more checking. AI panopticon.

Jan Kulveit avatar
⟳ AlignmentJournal retweeted
Jan Kulveit@jankulveit·Oct 23

Happy to be an editor: please do submit things I'd like to read. It is an ambitious attempt to fix academic publishing in this extremely important domain.

The Alignment Journal publishes research on the scientific foundations of understanding, predicting, and steering artificial intelligent systems. As of today, submissions are open to everyone. What we do differently: - We pay reviewers, with bonuses for speed and insight. - Review discussion is confidential, but every accepted paper is published with a “review summary” to distill the insights generated. - Each paper is judged against our acceptance bar, not against other submissions. - Papers under review are listed publicly, and researchers can nominate themselves to review, so each paper can reach the reviewers who know the topic best. We seek work that builds theoretical or otherwise principled analyses of agency, robustness, incentives, interpretability, generalization, and long-term behavior, and whose significance survives the deletion of any particular model, dataset or benchmark. Empirical investigations of such ideas are welcome, and negative results are encouraged. We are especially interested in approaches that could apply to future superintelligent systems that interact and self-improve. You can submit your preprint in any format. Submissions are rolling and open now on the website. If you are instead interested in reviewing or serving as an action editor (basically: an area chair), create an account on our website. Links to submit, our scope and the reasoning behind the Journal are in the reply ⬇️

Post attachment