Micah Benson @micahben.bsky.social · 10/06/2026See the website for more info: nemiconf.github.io/summer26/ Registration: forms.gle/qUNq84pB6AyU... Submission: forms.gle/PEfMyL4J3PL9...nemiconf.github.ioThe 3rd New England Mechanistic Interpretability (NEMI) Workshop 030
Micah Benson @micahben.bsky.social · 10/06/2026🧠🤖 The 2026 New England Mechanistic Interpretability (NEMI) Workshop will be Aug. 14 at Boston University! Help spread the word and join the New England mech interp community! Registration and submission info in thread:👇 1102
Reposted by Micah BensonWillie Agnew @willie-agnew.bsky.social · 26/03/2026One of the most common features of AI delusional spirals in our recent study is a belief that the AI is sentient or has a personality. This played a central role in the delusional narratives, and correlated with increased used. Regulators and AI developers should curb this! arxiv.org/abs/2603.16567arxiv.orgCharacterizing Delusional Spirals through Human-LLM Chat LogsAs large language models (LLMs) have proliferated, disturbing anecdotal reports of negative psychological effects, such as delusions, self-harm, and ``AI psychosis,'' have emerged in global media and… 172
Micah Benson @micahben.bsky.social · 25/03/2026I’m just a little PhD student so I’m still in the “this is awesome I can cold email anyone who’s research is cool” phase lol 221
Micah Benson @micahben.bsky.social · 25/03/2026Crazy I thought this was definitely about the US seeing it out of context… if Canada has fallen we are in deep 290
Micah Benson @micahben.bsky.social · 25/03/2026I truly believe the rapid advances in the mech interp subfield have something real to offer AI ethics researchers: A chance to look beyond the HOW of evals to the WHY, a first pass at a technical solution when we see the opportunity, a new avenue for showing failures that prove models are not gods 173
Reposted by Micah BensonWillie Agnew @willie-agnew.bsky.social · 25/03/2026There's a lot of external pressure on AI ethics to produce solutions instead of critique. As someone who's worked a lot on CSAM, NCII, mental health, and creative harms of AI, if AI developers would have only listened to critiques, we could have avoided all these harms in the first place. 2389
Micah Benson @micahben.bsky.social · 25/03/2026I ofc love trying to think of technical solutions, but find techno-solutionism is often asking the wrong Q. It’s the difference between: “Can we ensure an LLM therapist never encourages suicide?” vs. “Should we even try to use LLMs as therapists in the first place?” 000
Micah Benson @micahben.bsky.social · 24/03/2026Wow love this, going to make getting one of those a goal of mine 000
Micah Benson @micahben.bsky.social · 24/03/2026Why XAI 😭?? You gotta stage an intervention next time 010
Micah Benson @micahben.bsky.social · 24/03/2026How could we restructure academic incentives to reward policy work? 120
Micah Benson @micahben.bsky.social · 24/03/2026It’s pretty cool that they named this platform after a Wilco album 050
