🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

It was great to work with Amartya Sanyal, Tim Rocktäschel, and Edward Grefenstette at FAIR London. This line of research is fascinating! Thank you for the opportunity! Additional gratitude to Roberto Calandra for the support and advice.

Excited to announce that our #AAMAS2022 paper "Lazy-MDPs: Towards Interpretable RL by Learning When to Act" is on arXiv! 🦥 tl;dr - we introduce lazy-MDPs, modified MDPs that allow agents to defer decision-making to a third-party policy 📜 arxiv.org/abs/2203.08542 🧵👇

A common belief is that text auto encoders produce badly structured latent spaces with holes. We were surprised to find that using round-trip translations (e.g. en->de->en) one can obtain nicely structured latent spaces. Check out arxiv.org/pdf/2209.06792….

Very proud to contribute to making RL agents more accessible and reproducible!

We* are looking for a Student Researcher** to work with us on a project in intersection of modeling/generating speech/audio, NLP, and representation learning. *AudioLM team @ Google Research (Zalán Borsos, Neil Zeghidour, myself and many others!) **not-last-year PhD student

Our #ACL2023 paper "Factually Consistent Summarization via Reinforcement Learning with Textual Entailment Feedback" is now on arXiv! tl;dr - we improve the factuality of summaries via RL, without human feedback! 📜 arxiv.org/abs/2306.00186 Thread (1/10) 👇

Google presents MusicRL Aligning Music Generation to Human Preferences paper page: huggingface.co/papers/2402.04… propose MusicRL, the first music generation system finetuned from human feedback. Appreciation of text-to-music models is particularly subjective since the concept of

Very proud of the work done by Geoffrey Cideron , one of my last projects at Google. When we released MusicLM in May ’23, we incorporated a feedback system to realize the first ever large-scale, organic improvement of music generation through RLHF. 🎶🧵

Online feedback is crucial for alignment, so we propose a simple recipe to make any direct alignment method (think DPO / IPO / SLiC-HF) online using AI feedback 🧙‍♂️ In human evals, online methods yield on avg 66% wins, 28% ties and 6% losses vs offline methods (on TL;DR) 👀

I am so proud to see Gemma released today! I have had a fantastic time working on post-training and RLHF with an amazing team. Cannot wait to see what the community builds with these models!

I am very happy to announce that Gemma 1.1 Instruct 2B and “7B” are out! Here are a few details about the new models: 1/11