Summer Readings: AI Papers Showcase | Spring Labs x AISO x SAIN
Summer Readings: AI Papers Showcase
The classroom is: NU-4A06 (fourth floor, wing A)
Seven papers, seven short talks: a summer reading list built by curious minds like you.
We're closing the (AI) spring season the way you'd want to open a summer: with a reading list. Over one evening, seven AI papers get the rapid-fire treatment: short talks, big ideas, just enough to send you off wanting to read the full thing.
Want to present a paper?
Submit a proposal right here.
If your proposal is selected, the stage is yours: a 5-to-10-minute slot to present it to a room of fellow students, researchers, and curious minds.
Who is this for?
Everyone's welcome to come, listen, and ask questions. Each talk is meant to give a high-level feel for a paper and spark enough curiosity to read it in full. A background in AI helps, but it isn't needed, and we won't go deep into the technical weeds. AI professors will be in the room too, helping steer the discussions.
The topics
We have selected the readings from the following four tracks, revolving around frontier AI research.
(1) 🛡️AI Safety: alignment, interpretability, and the hard problems of responsible development
Väinö Bezuijen
Paper: The Alignment Problem from a Deep Learning Perspective: https://arxiv.org/pdf/2209.00626
Maya Ozbayoglu
Paper: What Drives Representation Steering? A Mechanistic Case Study on Steering Refusal: https://arxiv.org/pdf/2604.08524
(2) 🤖Foundation: landmark papers shaping the field
Steven Wagter
Paper: Think Fast: Estimating No-CoT Task-Completion Time Horizons of Frontier AI Models: https://arxiv.org/pdf/2606.07157
Duje Vukovac
Paper: Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models
arxiv.org/pdf/2503.16419
(3) 🃏Free Topic (extra safety paper)
Marysia Winkels
Paper: How Vulnerable Are AI Agents to Indirect Prompt Injections? Insights from a Large-Scale Public Competition: https://arxiv.org/abs/2603.15714
(4) 🧑🎨State of the Art: cutting-edge methods pushing what's possible
Rustam Eynaliyev
Paper: Grounding Computer Use Agents on Human Demonstrations: https://arxiv.org/abs/2511.07332
⁉️ Surprise Presentation on Continual Learning
we are still arranging the last presentation. Therefore, it will be a surprise.
Agenda
17:30 – 17:45 Introductions: settle in and get to know the room while we wait for the latecomers
17:45 – 19:15 Presentations: seven talks, hard-capped at ten minutes, each followed by a few questions
19:15 – onwards Drinks & snacks (on the house)
🎟️ Registration
Free for everyone: just register here on Luma to save your spot.
About us
This event is supported by AISO and SAIN, among the largest AI grassroots organisations in the Netherlands. It's conceived and run by Spring Labs, a fledgling student-led community focused on foundational AI research.
The space and catering is funded by Vrije Universiteit Amsterdam.
See you there!