Daniil Tiapkin (@dtiapkin) header
Daniil Tiapkin (@dtiapkin) profile picture

Research Scientist @ Google DeepMind | PhD in RL 🇫🇷

Paris, France
Joined June 2022
Photo by Daniil Tiapkin (@dtiapkin)Photo by Daniil Tiapkin (@dtiapkin)Photo by Daniil Tiapkin (@dtiapkin)Photo by Daniil Tiapkin (@dtiapkin)Photo by Daniil Tiapkin (@dtiapkin)Photo by Daniil Tiapkin (@dtiapkin)Photo by Daniil Tiapkin (@dtiapkin)
Daniil Tiapkin retweeted
Happy to share we just published Gemma 4 technical report! Take a look
37
125
10
1,058
91,561
Daniil Tiapkin retweeted
Excited to announce that our new book Reinforcement Learning Foundations is available for preorder! W/ @shiemannor and @YishayMansour 30% discount and free shipping till July 21 Discount code 148294 cambridge.org/il/universityp…
3
44
279
15,139
Daniil Tiapkin retweeted
Meet DiffusionGemma! An experimental open model that explores a fast approach to text generation, released under an Apache 2.0 license. Moving beyond sequential, token-by-token processes to generate entire blocks of text simultaneously. Here’s what’s new with DiffusionGemma: 👇
161
794
273
4,944
1,006,161
Daniil Tiapkin retweeted
Happy to share our recent preprint: Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation Takeaways: - standard implementations of uniform diffusion do not learn a denoiser, but rather a "leave-one-out" denoiser - uniform diffusion and masked diffusion are two sides of the same coin paper: arxiv.org/abs/2605.22765 code: github.com/samsongourevitch/… (we release 9 models trained on owt) with @samsongvch @dario_sha @umutsimsekli @EricMoulines @ericxing @AlainDurmus A short thread on our findings: (1/5)
4
25
4
107
10,979
🚀COSMOS is OUT! @NeurIPSconf 2025! 📈COSMOS achieves up to 2x faster text generation compared to other diffusion models, utilizing up to 8x compression in text representations for superior efficiency. 📄 Paper: arxiv.org/pdf/2506.21170 💻 Code: github.com/MeshchaninovViach… (1/6)
1
4
1
11
1,250
While frontier labs are announcing their new models, we also want to be part of this parade. So, we’re happy to announce gfnx – a JAX-first library with environments and a single-file baseline implementation for GFlowNet research.
4
10
21
1,463
Environments, reward functions, metrics, and single-file implementations – everything you need to achieve up to 80× single-seed speedups for combinatorial object generation, from bit sequences and Ising models to phylogenetic trees.
2
7
333
Daniil Tiapkin retweeted
1/ Can we efficiently learn the destruction process of diffusion samplers? Can we learn not just the drift, but also the variance for all transition kernels? – We answer YES in our recent paper “Adaptive Destruction Processes for Diffusion Samplers” (Oral at NeurIPS 2025 FPI Workshop).
1
9
1
17
2,805
The speedrun is over: I defended my PhD this week and became a doctor in applied mathematics (unofficially: in reinforcement learning)! Huge thanks to my supervisors (Eric & Gilles), collaborators, and friends for all the support.
6
2
2
30
3,249
If you're at #ICML2025, come say hi and learn about teacher hacking in distillation. See you at poster E-2706!
1/ If you’re familiar with RLHF, you likely heard of reward hacking —where over-optimizing the imperfect reward model leads to unintended behaviors. But what about teacher hacking in knowledge distillation: can the teacher be hacked, like rewards in RLHF?
1
12
649
Daniil Tiapkin retweeted
(1/n) The usual assumption in GFlowNet environments is acyclicity. Have you ever wondered if it can be relaxed? Does the existing GFlowNet theory translate to the non-acyclic case? Is efficient training possible? We shed new light on these questions in our latest work! @icmlconf
1
6
1
13
2,081
I'll be at #ICLR2025 this week - let's chat about RL, sampling and so on! Excited for @gritsaev to present our work on backward policy optimization for GFlowNets (arxiv.org/abs/2410.15474, my first work as advisor!) on Sat morning, April 26, poster 454. Come to say hi to us!
4
15
446
Daniil Tiapkin retweeted
1/ GFlowNets are known for training a forward policy to generate complex objects step by step. However, an equally important piece specific to the GFlowNet paradigm is a backward policy, which undoes these steps and plays a crucial role in training.
1
3
2
12
1,591
1/ If you’re familiar with RLHF, you likely heard of reward hacking —where over-optimizing the imperfect reward model leads to unintended behaviors. But what about teacher hacking in knowledge distillation: can the teacher be hacked, like rewards in RLHF?
3
13
3
43
15,382
5/ Our suggestions are the following: - Use online generations during distillation; - Train on more diverse prompt datasets; - Expand the dataset with multiple completions per prompt.
2
5
452