🇰🇷 ICML 2026

3 Papers at ICML next week!

Come find me if you want to chat about anthropomorphic misalignment, AI safety & governance, student mentorship, or research collaboration.
I am also joining the FAR AI workshop (hosting the opening panel), then the OpenAI Safety Mixer and AI Governance Afterparty.

Paper 1 · Position Paper 🎤 Oral
Anthropomorphic Misalignment Research Needs Stronger Evidence
Vansh Gupta* · Peter Nutter* · Samuel Stante* · Andreas Krause · Florian Tramer · Lukas Fluri† · Xin Chen† · Anna Hedström†
📄 Paper
🎙️ Oral  ·  Tue Jul 7  ·  1:30–1:45 PM  ·  Hall D2
🪧 Poster  ·  Wed Jul 8  ·  10:30–12:15  ·  Hall A
Anthropomorphic Misalignment Framework
Paper 2 · Workshops FAGEN & AIWild
Learning to Inject: Automated Prompt Injection via Reinforcement Learning
Xin Chen · Jie Zhang · Florian Tramèr
📄 Paper 💻 Code
🤖 FAGEN  ·  Fri Jul 10  ·  10:10–11:00 & 14:40–15:30  ·  Grand Ballroom 104–105
🛡️ AIWild  ·  Sat Jul 11  ·  10:40–12:00 & 16:10–17:00  ·  Hall B2
RLPI Pipeline RLPI Results
Paper 3 · Workshop CompLearn
Preference Instability in Reward Models: Detection and Mitigation via Sparse Autoencoders
Shunchang Liu · Xin Chen · Belén Martín-Urcelay · Francesco Croce
📄 Paper 💻 Code
🧠 CompLearn  ·  Sat Jul 11  ·  11:30–12:30  ·  Auditorium
Preference Instability Framework
Come say hi! 👋
@XinCynthiaChen  ·  ICML 2026  ·  Seoul, Korea