Dobrik Georgiev (@DobrikG)

Pinned post

Query-based KG RAG is finally SOTA. 🚀

The results:
📈 +16-23% gains
⚡ Up to 167x faster processing
🧠 Inductive (works with unseen graphs/relations)
🎯 Zero-shot

A joint work with our factuality team within @GCResearchTeam
(1/X)🧵

Dobrik Georgiev @DobrikG Oct 2

Honoured to receive the "John Atanasoff" Award from the President of Bulgaria, Iliyana Yotova, for our research in machine learning and neural algorithmic reasoning!

Huge thanks to my incredible collaborators, mentors, and family for the support. Onwards to the next challenge!

Dobrik Georgiev @DobrikG May 13

If you know me, perhaps the fact that even ICML recognised my reviewing skills should not surprise you.

I'd like to take a moment, however, to rant that our reviews were of questionable quality

#ICML #Reviewing

Dobrik Georgiev @DobrikG Nov 19, 2025

We are looking for someone for the Cambridge office too...

And I particularly am looking for someone to throw crazy ideas at🤪!

Graphcore Research Verified account @GCResearchTeam Nov 19, 2025

🚨 Graphcore is hiring AI Research Interns! 🚨

Join us to work at the intersection of hardware and AI and help shape the future of AI systems. Whether you're excited about efficient inference, large-scale training, or advancing frontier-model capabilities, we’ve got cutting-edge projects for you to dive into.

Interested? Apply below 👇
View quoted post on X

Dobrik Georgiev @DobrikG Oct 10, 2025

Reviewer #2 in me continues to live on! Read our rants on some very recent research 👇

Graphcore Research Verified account @GCResearchTeam Oct 10, 2025

September's Papers of the Month is here, and this month is all about LLMs! 🧠

Out of all papers released this month, our editor @robhu92 has curated:

📊 "FlowRL: Matching Reward Distributions for LLM Reasoning“ (review by @samot_gc): A clever usage of #GFlowNets to align an #RL policy model with the *full* reward distribution, encouraging mode coverage

☁️ "Soft Tokens, Hard Truths" (review by @lukehudlass): A simple and scalable method to integrate #continuousthoughts with RL training schemes

🏎️ "Set Block Decoding is a Language Model Inference Accelerator" (review by @douglasahorr): Exactly what it says on the 🥫!

🔁 "Turning Recurring LLM Reasoning into Concise Behaviors" (review by @DobrikG): An improvement to #CoT reasoning, trying to reduce self-repetition via metacognitive reduce

Check out our reviews in the 🧵 below!

View quoted post on X

Dobrik Georgiev

Latest posts 5