Luyang Liu

Research Scientist @ Google DeepMind

[Google DeepMind @GoogleDeepMind]

This is Gemini 3: our most intelligent model that helps you learn, build and plan anything.

It comes with state-of-the-art reasoning capabilities, world-leading multimodal understanding, and enables new agentic coding experiences. 🧵

[Luyang Liu @luyang1125]

🚀🚀🚀

[Demis Hassabis @demishassabis]

It's nearly 3 here, my favourite part of the night shift… locked in... 💪🚀

[Google DeepMind @GoogleDeepMind]

Scale your production apps with the stable version of Gemini 2.5 Flash-Lite. ⚡

It’s faster than our 2.0 Flash models, more cost-efficient, and outperforms 2.0 Flash-Lite across coding, math, reasoning, and multimodal understanding.

Start building → https://t.co/qtYRn9192A

[Luyang Liu @luyang1125]

Gemini Deep Thinking achieved gold-medal level in IMO, verified by the IMO organizers!

[Thang Luong @lmthang]

Very excited to share that an advanced version of Gemini Deep Think is the first to have achieved gold-medal level in the International Mathematical Olympiad! 🏆, solving five out of six problems perfectly, as verified by the IMO organizers! It’s been a wild run to lead this effort and I am grateful to everyone in the team for such an amazing achievement! Blog post in the thread and more to share soon!

[Thang Luong @lmthang]

Yes, there is an official marking guideline from the IMO organizers which is not available externally. Without the evaluation based on that guideline, no medal claim can be made. With one point deducted, it is a Silver, not Gold.

[Luyang Liu @luyang1125]

Very excited to see Gemini Diffusion I’ve been working on lately go live!

[Google DeepMind @GoogleDeepMind]

We’ve developed Gemini Diffusion: our state-of-the-art text diffusion model.

Instead of predicting text directly, it learns to generate outputs by refining noise, step-by-step. This helps it excel at coding and math, where it can iterate over solutions quickly. #GoogleIO

[Marktechpost AI @Marktechpost]

Google DeepMind Introduces Differentiable Cache Augmentation: A Coprocessor-Enhanced Approach to Boost LLM Reasoning and Efficiency

Researchers from Google DeepMind have introduced a method called Differentiable Cache Augmentation. This technique uses a trained coprocessor to augment the LLM’s key-value (kv) cache with latent embeddings, enriching the model’s internal memory. The key innovation lies in keeping the base LLM frozen while training the coprocessor, which operates asynchronously. The researchers designed this method to enhance reasoning capabilities without increasing the computational burden during task execution.

The methodology revolves around a three-stage process. First, the frozen LLM generates a kv-cache from an input sequence, encapsulating its internal representation. This kv-cache is passed to the coprocessor, which processes it with additional trainable soft tokens. Not tied to specific words, these tokens act as abstract prompts for generating latent embeddings. Once processed, the augmented kv-cache is fed back into the LLM, enabling it to generate contextually enriched outputs. This asynchronous operation ensures the coprocessor’s enhancements are applied efficiently without delaying the LLM’s primary functions. Training the coprocessor is conducted using a language modeling loss, focusing solely on its parameters while preserving the integrity of the frozen LLM. This targeted approach allows for scalable and effective optimization.....

Read the full article here: https://t.co/YpX9LbxOfh

Paper: https://t.co/4OELMxdyBl

@PfeiffJo @pepollopep @luyang1125 #ArtificialIntelligence @GoogleDeepMind

[Jonas Pfeiffer @PfeiffJo]

🧠💡 Our LLMs just had a ‘memory augmentation’—now they can deliberate like seasoned thinkers!

https://t.co/Czi7JyQTlr