Mastodawn

🌟 𝐈𝐧𝐭𝐫𝐨𝐝𝐮𝐜𝐢𝐧𝐠 [Google Research’s] 𝐆𝐫𝐨𝐮𝐧𝐝𝐬𝐨𝐮𝐫𝐜𝐞 - 𝐀𝐧 𝐨𝐩𝐞𝐧 𝐬𝐨𝐮𝐫𝐜𝐞 𝐝𝐚𝐭𝐚𝐬𝐞𝐭 𝐨𝐟 𝐡𝐢𝐬𝐭𝐨𝐫𝐢𝐜 𝐟𝐥𝐨𝐨𝐝 𝐞𝐯𝐞𝐧𝐭𝐬 𝐟𝐫𝐨𝐦 𝐧𝐞𝐰𝐬 𝐚𝐫𝐭𝐢𝐜𝐥𝐞𝐬.
--
https://doi.org/10.31223/X5RR2K / https://eartharxiv.org/repository/view/12083/ <-- shared paper
--
https://zenodo.org/records/18647054 <-- shared link to associated dataset
--
https://sites.research.google/gr/floodforecasting/ <-- shared link to Google Research flood forecasting effort entry page
--
#GoogleResearch #Google #Gemini #AI #ClimateTech #MachineLearning #DataScience #FloodForecasting #Sustainability #TechForGood #aggregation #curated #newsarticles #news #media #article #harvesting #reports #reporting #global #world #historic #naturalhazards #naturaldisaster #floods #flooding #flashflood #water #hydrology #extremeweather #climatechange #GDACS #Groundsource #GIS #spatial #mapping #spatialanalysis #spatiotemporal #geographic #openaccess #openscience #opendata #floodevents #LLM #gemini #largelanguagemodel #deeplearning #AI #precision #metrics #historicresource #model #modeling #forecasting #opensource

sayzard 6d ago

fly51fly (@fly51fly)

Google Research와 University of Wisconsin-Madison 공동 저자들의 2026년 논문 'Grow, Don't Overwrite: Fine-tuning Without Forgetting'(arXiv)이 공유되었습니다. 미세조정(fine-tuning) 시 기존 지식을 잃지 않도록 하는 방법을 제안하는 연구로, 모델 지속 학습과 캐타스트로픽 포게팅 문제에 대한 해결책을 다룹니다. arXiv 링크 포함.

https://x.com/fly51fly/status/2031486326000988505

#finetuning #continuallearning #googleresearch #arxiv

fly51fly (@fly51fly) on X

[LG] Grow, Don't Overwrite: Fine-tuning Without Forgetting D Adila, H Mazzawi, B Dherin, X Gonzalvo [Google Research & University of Wisconsin-Madison] (2026) https://t.co/gTyQoHNgUU

X (formerly Twitter)

sayzard Feb 24

[프롬프트 반복으로 LLM 정확도 향상, Google 팀 연구 결과

Google Research 팀이 발표한 논문에서, 프롬프트를 두 번 반복 입력하면 대부분의 최신 LLM(Gemini, GPT-4o, Claude, DeepSeek 등)의 정확도가 크게 향상된다는 결과를 발견했습니다. 이 기법은 LLM의 인과적 구조의 취약점을 보완하며, 특히 단순 참조, 목록 처리, 구조화된 데이터 질문에 효과적입니다.

https://news.hada.io/topic?id=26960

#llm #promptengineering #googleresearch #gemini #gpt4

프롬프트 반복으로 LLM 정확도 향상, Google 팀 연구 결과

<p>Google Research 팀이 발표한 논문(“Prompt Repetition Improves Non-Reasoning LLMs”)에서 발견한 매우 간단하면서도 강력한 기법:<br /> <str...

GeekNews

AI Sparkup Feb 24

프롬프트 두 번 붙여넣기로 LLM 정확도 높이는 방법, Google 연구 결과

Google Research가 발견한 프롬프트 반복 기법. LLM에 같은 프롬프트를 두 번 입력하면 비용·지연 증가 없이 정확도가 오릅니다.

https://aisparkup.com/posts/9574

Show thread

strickvl Feb 12

https://research.google/blog/scheduling-in-a-changing-world-maximizing-throughput-with-time-varying-capacity/

#GPUScheduling #MLInfra #GoogleResearch

Scheduling in a changing world: Maximizing throughput with time-varying capacity

sayzard Feb 11

Google DeepMind (@GoogleDeepMind)

Google Research와 공동으로 발표된 두 편의 논문에서 'Gemini Deep Think'가 에이전트형(agentic) 워크플로를 활용해 수학·물리학·컴퓨터과학 수준의 연구 문제를 해결하는 방법을 제시합니다. AI를 연구 협업자로 활용하는 새로운 접근과 사례를 다룬 연구 결과입니다.

https://x.com/GoogleDeepMind/status/2021632302070026581

#gemini #googleresearch #agentic #researchcollab

Google DeepMind (@GoogleDeepMind) on X

How could AI act as a better research collaborator? 🧑‍🔬 In two new papers with @GoogleResearch, we show how Gemini Deep Think uses agentic workflows to help solve research-level problems in mathematics, physics, and computer science. More → https://t.co/ISWedYjpgD

X (formerly Twitter)

ResearchBuzz: Firehose Jan 31

Maps Mania: Meet Language Explorer: Google’s New Open-Source Linguistic Atlas. “Google Research’s Language Explorer, is a new interactive map that anyone can use to explore the world’s languages.”

https://rbfirehose.com/2026/01/31/meet-language-explorer-googles-new-open-source-linguistic-atlas-maps-mania/

Meet Language Explorer: Google’s New Open-Source Linguistic Atlas (Maps Mania)

Maps Mania: Meet Language Explorer: Google’s New Open-Source Linguistic Atlas. “Google Research’s Language Explorer, is a new interactive map that anyone can use to explore the world&#8…

ResearchBuzz: Firehose

sayzard Jan 13

Chubby (@kimmonismus)

Google Research가 MedGemma 1.5를 공개해 CT·MRI·병리조직·X선 등 영상 및 의료 텍스트 전반에서 성능을 크게 향상시켰다고 보고했습니다(이미지 작업 최대 +14%, EHR QA 최대 +22%). 발표와 함께 MedASR(오픈 의료 관련 도구)도 함께 언급되었습니다. 의료 AI 성능 개선의 중요한 업데이트입니다.

https://x.com/kimmonismus/status/2011184297026179305

#medgemma #medicalai #googleresearch #medasr #healthcareai

Chubby♨️ (@kimmonismus) on X

Google Research has released MedGemma 1.5, a major upgrade that boosts performance across CT, MRI, histopathology, X-ray timelines, and medical text Delivering gains up to +14% accuracy in imaging tasks and +22% in EHR QA. Alongside it comes MedASR, an open medical

X (formerly Twitter)

techUpdate.io Nov 18

WeatherNext 2: Achtmal schnelleres Wetter - Weil Zeit auch bei Wolken Geld ist

https://techupdate.io/google/weathernext-2-achtmal-schnelleres-wetter-weil-zeit-auch-bei-wolken-geld-ist/50671/

#technews #weathernext2 #deepmind #kiwetter #googleresearch

WeatherNext 2: Achtmal schnelleres Wetter - Weil Zeit auch bei Wolken Geld ist

Google stellt WeatherNext 2 vor: Das neue KI-Wettermodell von DeepMind liefert achtmal schnellere, stundengenaue Prognosen.

techUpdate.io

Saarland Informatics Campus Nov 7

🏆 Congrats to Yonggang Jiang, PhD candidate at the Saarbrücken Graduate School of Computer Science (Saarland University) and researcher at the Max Planck Institute for Informatics, on being named a 2025 Google PhD Fellow!

🔗 More about his research: https://sic.link/14

@mpiinf @DFKI

#SIC #SaarlandUniversity #MPIInformatics #Algorithms #Theory #GoogleResearch