Marius Mosbach
@mariusmosbach
NLP postdoc at Mila & McGill 🇨🇦 🦋 http://mariusmosbach.bsky.social
I am looking for candidates interested in pursuing a Postdoc position in my lab @Mila_Quebec. Catch me at #NeurIPS2024 or ping me on the Whova App if you are interested. My lab's focus is on multimodal vision-language research. Below are some example research topics that are of…
Very excited to have Abhilasha in Germany 🇩🇪 and very close to where I'm from 😅! If you are interested in doing a PhD, you should definitely reach out to her!
Life update: I’m excited to share that I’ll be starting as faculty at the Max Planck Institute for Software Systems(@mpi_sws_) this Fall!🎉 I’ll be recruiting PhD students in the upcoming cycle, as well as research interns throughout the year: lasharavichander.github.io/contact.html
Life update: I’m excited to share that I’ll be starting as faculty at the Max Planck Institute for Software Systems(@mpi_sws_) this Fall!🎉 I’ll be recruiting PhD students in the upcoming cycle, as well as research interns throughout the year: lasharavichander.github.io/contact.html
Hope everyone’s getting the most out of #icml25. We’re excited and ready for the Actionable Interpretability (@ActInterp) workshop this Saturday! Check out the schedule and join us to discuss how we can move interpretability toward more practical impact.
Mechanistic interpretability often relies on *interventions* to study how DNNs work. Are these interventions enough to guarantee the features we find are not spurious? No!⚠️ In our new paper, we show many mech int methods implicitly rely on the linear representation hypothesis🧵
I'll be at #ICML2025 this week presenting SafeArena (Wednesday 11AM - 1:30PM in East Exhibition Hall E-701). Come by to chat with me about web agent safety (or anything else safety-related)!
🚨Meet our panelists at the Actionable Interpretability Workshop @ActInterp at @icmlconf! Join us July 19 at 4pm for a panel on making interpretability research actionable, its challenges, and how the community can drive greater impact. @nsaphra @saprmarks @kylelostat @FazlBarez
Have you ever wondered whether a few times of data contamination really lead to benchmark overfitting?🤔 Then our latest paper about the effect of data contamination on LLM evals might be for you!🚀 "How Much Can We Forget about Data Contamination?" (accepted at #ICML2025) shows…
Going to #icml2025? Don't miss the Actionable Interpretability Workshop (@ActInterp)! We've got an amazing lineup of speakers, panelists, and papers, all focused on leveraging insights from interpretability research to tackle practical, real-world problems ✨
🚨Excited to release OS-Harm! 🚨 The safety of computer use agents has been largely overlooked. We created a new safety benchmark based on OSWorld for measuring 3 broad categories of harm: 1. deliberate user misuse, 2. prompt injections, 3. model misbehavior.
if you have no other joy in your life other than thinking of your importance as AI researcher then GO FIND SOMETHING ELSE (e.g. sports, arts, friends, kids) stay grounded and preserve your mental health this may sound trivial, but given recent events, worth repeating
Who's gonna proof that finding a problem in NLP that nobody else is working on is also NP-complete?
2️⃣ concurrent preprints proving tokenization is NP complete!
📣Happy to (pre-)release my Fleurs-SLU benchmark to evaluate massively multilingual spoken language understanding on SIB & Belebele. Work done at @Mila_Quebec with @davlanade @gg42554 @licwu Datasets: huggingface.co/datasets/WueNL… huggingface.co/datasets/WueNL… Details to follow👇
Come by tomorrow 11am-2pm at @NeurIPSConf Poster #1606 to chat more about AURORA 🌌, text-guided editing, and why it is arguably more interesting than image generation Or anything related to world models, evals/analysis/interp, vision+language reasoning, cogsci, academic life!
AURORA 🌌 is now accepted as a Spotlight at NeurIPS 🥂 We wondered if a model can do *controlled* video generation but in a *single* step? So we built a dataset+model for “taking actions” on images via editing, or what you could call single-step controlled video gen
Excited to be at #NeurIPS2024 this week. Happy to meet up and chat about retrievers, RAG, embedders etc, or anything LLM-related!
I am looking for a postdoc at Mila/McGill. Catch me at #NeurIPS2024 if interested. Perks: . pursue own interests . mentor talented PhDs . collaborate with Mila profs and industry . access to large compute . live in a vibrant city
I will be at #NeurIPS2024 Wed and Thu. Tomorrow at UBC for the Future of NLP event presenting "Learning to reason with Generative Models", covering post-training methods and inference time reasoning for LLMs and vision (diffusion) models. Happy to meet anyone interested!
I will be at #NeurIPS2024 and the Future of NLP Workshop at @UBC_NLP next week to present our LLM2Vec poster 🤩🔥 Let’s chat if you are interested ☺️ arxiv.org/abs/2404.05961
Exciting times for "doing actual RL" with Tulu3 and likely Llama 3.3! Heading to #NeurIPS to present VinePPO at MATHAI, where we built the right RL framework for LLM reasoning. Let’s chat about - RL training of LLMs - Learning to search - Test-time scaling DM me if interested!
What's the one advise you wish people would have told you before applying to a faculty position?