#Research
Curated coverage on «Research»: releases, research, benchmarks, open-source.
Google released SpeciesNet — an open-source AI model for wildlife protection
Google introduced SpeciesNet, an open-source model that helps conservationists around the world identify animal species from camera-trap images. How AI is becoming…
97 Hours on a Single GPU: An Experiment With a Self-Learning Neural Network and the Trap of Impressive Metrics
An independent researcher tried to teach a language model to plug in new skills like apps — without retraining. The architecture worked, the metrics looked perfect…
Reasoning Models Can't Hide Their Thoughts — And That's Good
OpenAI introduced the CoT-Control tool and discovered: reasoning models are nearly unable to control their chains of thought. Paradoxically, this very 'weakness'…
AI helped compute graviton amplitudes in quantum gravity
Edge AI in video cameras: how OpenIPC, an open project, is changing the rules of video analytics
Project Genie by Google DeepMind: how to create entire worlds with text prompts
Hugging Face trained an image generation model in 24 hours
Decima-8: the architecture seeking to reinvent neuromorphic chips
The new neuromorphic architecture Decima-8 proposes a radical approach: encoding signals with energy levels, digital crossbars instead of unreliable memristors, and…
AI helped formally verify a Fields Medal laureate’s proofs for the first time
The proofs of Ukrainian mathematician Maryna Viazovska on the sphere packing problem have undergone formal computer verification for the first time with AI…
Outpost Bio raised $3.5 million to build AI models of the human microbiome
Hugging Face engineers wrote a practical guide to generative AI: from transformers to fine-tuning
SHAP-IQ: a new standard for explainable AI enters practical use
Google introduces STATIC: 948x faster generative search
Goodbye, patches: TAPe + ML architecture changes the rules of computer vision
The T+ML architecture has been presented, using elements of active perception instead of raw pixels. This allows models to see object structure immediately, saving…
Unified Latents: Google DeepMind finds a way to improve AI generation
Google DeepMind's new Unified Latents framework solves the dilemma between data compression quality and neural network accuracy.