Athanasios Vlontzos, PhD

🤯 One book 10 pages. 100 TRILLION poems. And how it helps us understand LLMs a 🧵 🤯
Cent mille milliards de poèmes, is a real book by Raymond Queneau! It lets you create 100,000,000,000,000 different sonnets just by mixing and matching lines! 🤯 #LLM #AI #Poetry #Books #MindBlown
🧵 1/7

06.02.2025 14:16 — 👍 6    🔁 1    💬 1    📌 0

Excited to attend #NeurIPS2024 in Vancouver next week!

If you’re interested in chatting about Spotify’s work, causal ML, or the future of AI, let’s connect. Use this link to schedule a meeting: calendly.com/athanasiosvl....

Looking forward to great discussions and insights!

06.12.2024 14:31 — 👍 0    🔁 0    💬 0    📌 0

📢 5 days to @neuripsconf.bsky.social 2024! We're presenting our poster "Benchmarking Counterfactual Image Generation"!

🗓️ When: Wed, Dec 11, 11 a.m.–2 p.m. PST
📍 Where: West Ballroom A-D (#5008), Vancouver

Interested in counterfactual image generation or evaluation frameworks? Come chat! 🚀

06.12.2024 14:04 — 👍 2    🔁 1    💬 1    📌 0

Reach out if you are attending @neuripsconf.bsky.social next week to discuss the amazing work we do at @SpotifyResearch

02.12.2024 13:47 — 👍 0    🔁 0    💬 0    📌 0

We tackle a key challenge: estimating causal effects for complex treatments (text, images, etc.).

✅ Novel contrastive learning approach
✅ Unbiased causal estimates
✅ Outperforms on synthetic & real-world data

Implications for recsys, drug discovery & more!

#CausalInference #AI #MachineLearning

02.12.2024 13:45 — 👍 0    🔁 0    💬 1    📌 0

Contrastive representations of high-dimensional, structured treatments Estimating causal effects is vital for decision making. In standard causal effect estimation, treatments are usually binary- or continuous-valued. However, in many important real-world settings, treat...

🚀 Just dropped our latest research from @spotify Advanced Causal Inference Lab: “Contrastive Representations of High-Dimensional, Structured Treatments” 📜

arxiv.org/abs/2411.19245

02.12.2024 13:45 — 👍 5    🔁 1    💬 1    📌 0

Proud of our students for leading this effort! See you all in a couple weeks @neuripsconf.bsky.social !

26.11.2024 18:12 — 👍 0    🔁 0    💬 0    📌 0

Thank you to my co-authors @bernhardkainz.bsky.social and C. Müller

25.11.2024 13:50 — 👍 1    🔁 1    💬 0    📌 0

Thrilled to announce my contribution to a soon-to-be-released book on Trustworthy AI for Medical Imaging! 🚀I co-authored a chapter on causal reasoning with AI, exploring how it can enhance reliability & fairness in real-world clinical applications.
Stay tuned! 🧠📚#AI #Causality

25.11.2024 13:50 — 👍 3    🔁 0    💬 1    📌 0