Dear ChatGPT, Am I the Asshole?.While Reddit users might say yes, your favorite LLM probably won’t. We present Social Sycophancy: a new way to understand and measure sycophancy as how LLMs overly preserve users' self-image.

RT @stanfordnlp: Computer-vision research powers surveillance technology by @ria_kalluri, @chengmyra1, and colleagues out in @Nature!. “Her….

RT @jathansadowski: Very happy to see my invited article for Nature is out in the new issue. I wrote about computer vision, the AI surveill….

RT @irenetrampoline: What more could we understand about the fractal, “jagged” edges of AI system deployments if we had better ways to list….

made it to Athens for Facct!! reach out if you want to chat about perceptions of AI, metaphors, sycophancy, anthropomorphism, cats, or anything else :D #facctcats (he is reading about himself)

RT @ShirleyYXWu: Even the smartest LLMs can fail at basic multiturn communication. Ask for grocery help → without asking where you live….

RT @michaelryan207: New #ACL2025NLP Paper! 🎉. Curious what AI thinks about YOU?. We interact with AI every day, offering all kinds of feedb….

RT @baixuechunzi: Avoiding race talk can feel unbiased, but it often isn’t. This racial blindness can reinforce subtle bias in humans. Alig….

RT @shaily99: 🖋️ Curious how writing differs across (research) cultures?.🚩 Tired of “cultural” evals that don't consult people?. We engaged….

RT @lltjuatja: When it comes to text prediction, where does one LM outperform another? If you've ever worked on LM evals, you know this que….

mom pick me up im scared (exciting work!!).

RT @soldni: 2025 AI hot take: everyone should use FastText more. Word embeddings are awesome.

RT @manoelribeiro: This @acm_chi paper surveys 319 knowledge workers. The authors find that GenAI reduces cog load for some tasks (info sea….

@sunnyyuych @jurafsky Working on this has made me realize how weird and messy preference datasets are and how we can improve them (cc @enfleisig haha). If that's something you're also thinking about i'd love to chat!! or read your paper :D.

Paper: Code: Thanks to my wonderful collaborators @sunnyyuych @jurafsky and everyone who helped along the way!!.

So we built DumT, a method using DPO + HumT to steer models to be less human-like without hurting performance. Annotators preferred DumT outputs for being: 1) more informative and less wordy (no extra “Happy to help!”) 2) less deceptive and more authentic to LLMs’ capabilities.

We also develop metrics for implicit social perceptions in language, and find that human-like LLM outputs correlate with perceptions linked to harms: warmth and closeness (→ overreliance), and low status and femininity (→ harmful stereotypes).

First, we introduce HumT (Human-like Tone), a metric for how human-like a text is, based on relative LM probabilities. Measuring HumT across 5 preference datasets, we find that preferred outputs are consistently less human-like.

Do people actually like human-like LLMs? In our #ACL2025 paper HumT DumT, we find a kind of uncanny valley effect: users dislike LLM outputs that are *too human-like*. We thus develop methods to reduce human-likeness without sacrificing performance.

RT @stanfordnlp: Only ding a model for making mistakes! . It gives better results in RL and avoids mode collapse. We still understand so….

RT @ang3linawang: Have you ever felt that AI fairness was too strict, enforcing fairness when it didn’t seem necessary? How about too narro….