Exciting news! The WiCV@ICCV 2025 Dinner & Mentorship event will be held in Honolulu! 🗓️Oct 19, 2025 @ 7 PM RSVP now (spots are limited!): https://t.co/6ORdRWS2gF Only confirmed guests via email will be admitted. More info: https://t.co/8z7XTToWgL #WiCV #ICCV2025 #WomenInCV

It is PhD application season again 🍂 For those looking to do a PhD in AI, these are some useful resources 🤖: 1. Examples of statements of purpose (SOPs) for computer science PhD programs: https://t.co/Stz53ZiREM [1/4]

2 papers each accepted by 2 reviewers out of 3. No strong arguments against the paper, AC accepting the paper and the final result reject with no additional explanation? 😳 Any insights about @NeurIPSConf #NeurIPS2025 decision process this year?

"Build the web for agents, not agents for the web" This position paper argues that rather than forcing web agents to adapt to UIs designed for humans, we should develop a new interface optimized for web agents, which we call Agentic Web Interface (AWI).

Come join us!

Join: https://t.co/PezB5nHdRR More on the EvalEval Coalition https://t.co/8G4WcGabhB Or ask to understand more

🚀 Technical practitioners & grads — join to build an LLM evaluation hub! Infra Goals: 🔧 Share evaluation outputs & params 📊 Query results across experiments Perfect for 🧰 hands-on folks ready to build tools the whole community can use Join the EvalEval Coalition here 👇

I’ll be at my first @CVPR this week presenting our #Highlight paper on (hyperbolic) safety-awareness in VLMs along with my co-author @tobiapoppi! Friends on twitter, DM me if you are going too and let’s catch up to discuss research and other banter 🥳 #CVPR2025

Upon graduation, I paused to reflect on what my PhD had truly taught me. Was it just how to write papers, respond to brutal reviewer comments, and survive without much sleep? Or did it leave a deeper imprint on me — beyond the metrics and milestones? Turns out, it did. A

We ran tens of thousands of dollars of benchmarking experiments to see whether @vllm_project or @lmsysorg SGLang was "faster". The result: across the models and sequence lengths we tried, their performance is nearly identical, with no clear patterns. ¯\_(ツ)_/¯

Are you working on multilingual, multicultural #LLM? Interested in diverse & inclusive language modeling? 😎 Stay tuned at our MELT workshop collocated with #COLM2025 🔗 https://t.co/fHWfh5a1AQ 🫶 We welcome 2p (EA), 4p (short), 8p (long) papers as well as talented reviewers!

This work was done in collaboration with @mziizm @beyzaermis @stevebach and Julia Kreutzer, and took us nearly one year to complete it. 🙏 We are grateful for feedback at early stage from @HellinaNigatu @sbmaruf @CriMenghini @AlhamFikri @pjox13 @OjewaleV ​​@simi_97k

🧵 Multilingual safety training/eval is now standard practice, but a critical question remains: Is multilingual safety actually solved? Our new survey with @Cohere_Labs answers this and dives deep into: - Language gap in safety research - Future priority areas Thread 👇

This work was an amazing collaboration with Lauren(@nolauren), Maria Antoniak , Taylor Arnold, Jiaang(@jiaangli), Siddhesh(@whoSiddheshp), Antonia(@AntoniaKaramol), Stella Frank, Zhaochong( @ZhaochongAn), Negar(@negar_rz), Daniel(@daniel_hers),@SergeBelongie and @KatiaShutova

We find that decades of visual cultural studies offer powerful ways to decode cultural meaning in images!! Rather than proposing yet another benchmark, our goal was to revisit and re-contextualize foundational theories of culture to pave way for more inclusive frameworks.

We propose 5 frameworks to evaluate cultures in VLMs: 1⃣Processual Grounding-who defines culture 2️⃣Material Culture-what is represented? 3️⃣Symbolic Encoding-layered meaning 4️⃣Contextual Interpretation-who understands & frames meaning? 5️⃣Temporality-when is culture situated?

In this paper, we call for integrating methods from 3 fields: 📚 Cultural Studies – how values, beliefs & identities are shaped through cultural forms like images. 🔍 Semiotics – how signs & symbols convey meaning 🎨 Visual Studies – how visuals communicate across time & place

Modern Vision-Language Models (VLMs) often fail at cultural understanding. But culture isn’t just recognizing things like food, clothes, rituals etc. It's how meaning is made and understood; it also about symbolism, context, and how these things evolve over time.

New Preprint 🎉: "Cultural Evaluations of Vision-Language Models Have a Lot to Learn from Cultural Theory". We review recent works on culture in VLMs and argue for deeper grounding in cultural theory to enable more inclusive evaluations. Paper 🔗: https://t.co/9AoRHTFG58

Join us at CVPR 2025 for our workshop VLMs-4-All: Vision-Language Models for All! 🌍✨ We're tackling the challenge of building geo-diverse, culturally aware VLMs. If you're passionate about inclusivity in AI, we’d love your participation! #CVPR2025 #VLMs4All