New @Nature publication from @GoogleResearch & @GoogleDeepMind: In this study, we advance AMIE, our research medical AI, from one-off diagnostic conversations toward treating & managing disease over time, using clinical guidelines & drug formularies. More:

At #GoogleIO#, we announced Gemini for Science, a collection of new experimental tools developed by @GoogleResearch, @GoogleLabs, @GoogleDeepMind & @GoogleCloud. Learn more about how we’re supercharging scientific research:

How can we help scientists get better data out of their instruments? Using @GoogleDeepMind AlphaEvolve, @GoogleResearch identified new ways to train DeepConsensus, improving DNA sequencing accuracy and output on PacBio instruments. Read more:

Attending the Women in Machine Learning (WiML) Social at #ICLR2026#? Stop by 203C from 12PM - 3PM to connect with the community and the Google Research team! #WiML# #GoogleResearch#

Since introducing Empirical Research Assistance last fall, Google Research scientists have been using it to address real-world applications in epidemiology, cosmology, atmospheric monitoring, and neuroscience, providing a hint of AI’s transformational capabilities to accelerate scientific discoveries. Learn more →

🔍 You notice "red dots on your legs" — but could you search for "palpable purpura"? Google studied how far AI can help people actually understand skin conditions, with 2,345 participants. Title: Research into how AI can help users understand skin conditions URL: 💡 Overview More than half of adults search online for health info and one in three turn to AI. Google Research looked not at AI's diagnostic accuracy itself, but at how much AI actually helps laypeople make better decisions, across two user studies. 🩺 Challenges Solved People can spot "red dots" but rarely know to search "palpable purpura." Being able to access information is not the same as understanding it and choosing the right next step. This work tackles exactly that gap. 🧪 Methodology Using real de-identified cases with images and medical histories, 2,345 participants were randomized into three arms: a control using normal web search, an AI arm showing 3-7 image-backed predicted conditions, and a Wizard of Oz arm where the candidates were dermatologist-verified ground truth. A community study with 110 local participants tested a 4-language app in the real world. 📊 Experimental Results Condition-naming accuracy nearly tripled, 23% for AI versus 8% for control. In the community study, naming ability rose 260% and clinicians rated the app helpful 92% of the time. But accuracy on "see a doctor or self-care?" barely improved, and AI users tended to underestimate urgency. Identifying a condition is not the same as acting on it correctly. #AI# #HealthTech#

🤔 Even for simple factual questions that need no logic, making a model "think" (CoT) somehow raises accuracy. Google Research explains why. Title: Thinking to Recall: How Reasoning Unlocks Parametric Knowledge in LLMs URL: Reasoning is not just task decomposition; it is also a device for pulling knowledge out of the weights. Three highlights. 🧮 Computational buffer The reasoning tokens themselves act as extra room for latent internal processing. In fact, replacing a meaningful trace with meaningless repeated text ("Let me think") of the same length still improved recall over doing nothing, though it never matched natural reasoning and plateaued when stretched too far. 🔗 Factual priming Generating related facts first primes recall of the correct answer, much like spreading activation in human cognition. Conditioning on just the facts extracted from a trace recovered most of reasoning's gains and helped even with reasoning disabled. ⚠️ Fragile to hallucination The mechanism is a double-edged sword: a single hallucinated fact in the trace sharply reduces correct final answers. Conversely, simply selecting hallucination-free trajectories at test time considerably improved accuracy. The practical takeaway: training with process rewards that encourage factually-grounded intermediate steps could boost reliability and cut hallucination vulnerability. #LLM# #Reasoning#

That grainy tangle of colors is real brain tissue, mapped in more detail than any human brain sample ever has been. The sample came from a woman's temporal lobe, removed during surgery to treat her epilepsy, and donated for research afterward. A team from Google Research and Jeff Lichtman's lab at Harvard sliced it into roughly 5,000 wafer-thin sections, imaged each one with an electron microscope, then used machine learning to digitally stitch everything back into one continuous 3D reconstruction. No human could trace this by hand, the resulting dataset alone runs to 1.4 petabytes, more storage than most people will use in a lifetime. Inside that single cubic millimeter, roughly one millionth of an entire human brain, the team counted about 57,000 individual cells, nearly 150 million synapses connecting them, and 230 millimeters of blood vessels threading through the tissue. The colors in images like this one aren't decorative, they're how researchers visually separate individual neurons and cell types from the tangle around them so they can trace each one's path. The map turned up some genuine surprises too. Most neuron pairs connect through just a couple of synapses, but researchers found rare pairs linked by up to 50 connections at once. They also spotted neurons whose branches curl around and knot into themselves, something nobody had documented before, along with pairs of neurons that turned out to be near perfect mirror images of each other. Knowing this level of detail exists for just one millionth of a brain, how far off do you think a full human connectome actually is?