Kaggle

6,000+ submissions. 12 standout teams. 🏆 We're excited to announce the top 12 teams who showcased exceptional creativity, technical expertise, and problem-solving abilities through the power of AI agents. Thank you to the 350,000+ learners who joined the 5-Day AI Agents: Intensive Vibe Coding Course with Google, and to everyone behind the 6,000+ submissions that showcased innovation and collaboration across the globe. Check out the winning teams and their submissions here: https://lnkd.in/gaWpwxvN

The role of a Kaggle competitor is expanding from writing code to also directing agents. Megan Risdal sat down with Taha Bouhsine to discuss, amongst other topics: • How some Grandmasters are managing "agent teams"  • Why the agentic era made Kaggle more accessible • How understanding AI’s limits separates the winners 📽️ Watch the complete interview here: https://lnkd.in/ek7iA-fd 

We’ve added Gemini 3.6 Flash and 3.5 Flash-Lite to Kaggle Benchmarks! You can now stress-test how these faster, token-efficient models perform at scale within the Kaggle Benchmarks ecosystem, at no cost. Run your own benchmarks and see how they compare! Get started here: https://lnkd.in/eMFrztr2

Backing a machine learning challenge with a real reward builds a natural "accountability loop". When there is an incentive on the line, participant friction naturally keeps the objectives rigorous and motivation high. Watch this clip from the latest Kaggle Conversations, where Megan Risdal sits down with Sara Hooker, CEO and co-founder of adaption, former VP of Research at Cohere, and member of Kaggle's Research Advisory Board to break down competition dynamics and human incentives in AI. Enjoy the full conversation at https://lnkd.in/e7h_cvVd

🏆 Announcing the winners of the Measuring Progress Toward AGI Hackathon! Advancing AI requires better ways to understand what models can do, where they excel, and where they still fall short. Thank you to the 1,000+ teams that submitted benchmark ideas spanning five cognitive tracks, creating new ways to evaluate model capabilities and limitations. The four winners for the Grand Prizes ($25,000 each) are: 🏅 MEDLEY-BENCH by Farhad Abtahi, Abdolamir Karbalaie, Eduardo Illueca-Fernandez, and Fernando Seoane: Tests whether models can recognize when they're wrong and hold firm when socially pressured to change a correct answer. 🏅 LearningBench by Karandeep Singh: Measures whether models can learn the rules of entirely new systems from scratch within a single conversation. 🏅 GAUGE by Arjun Thilak and Ramkumar Muniandi Vallimayil: Tests whether models know when to submit answers vs. abstain when uncertain. 🏅 Metaproteus by Joseph Boskovski: Tests whether models accurately predict their own likely responses, revealing gaps in self-knowledge. Find the 4 Grand Prize and 10 Track Prize winners here: https://lnkd.in/gmzagArw.

In 2020, Sara Hooker published "The Hardware Lottery", exploring a reality that shapes AI progress today: thriving software ideas are often just the ones that fit currently available hardware best. In this episode of Kaggle Conversations, Megan Risdal sits down with Sara Hooker, CEO and co-founder of adaption, former VP of Research at Cohere, and member of Kaggle's Research Advisory Board to discuss her thesis, also diving into: • What’s after scaling: Why the Transformer architecture is saturating and shifting the advantage back to recipes and test-time exploration. • The “Leaderboard Illusion”: How the pressure to report success has led to distortion and manipulation in open AI evaluations. • The underrated area of AI machine learning: Why computer scientists need to start taking the interface more seriously. Watch an interview clip below and the full conversation here: https://lnkd.in/e7h_cvVd

Ready to build an AI Training Agent to battle in the Pokémon TCG? Check out this starter notebook provided by the Pokémon Company, a good place to begin with: • Construct competitive decks • Evaluate matchups against other agents • Submit a working AI to the competition battle Find the notebook here: https://lnkd.in/etu6aE_z There are two competitions currently live! Submit your best agent to the Simulation competition by August 16, 2026, and share your approach and reasoning in the Strategy competition by September 13, 2026. The Prize Pool for the Strategy Category is $240,000. Get started here 👉 https://lnkd.in/eqSFPAxk

How to approach benchmarks in a world where saturation is skyrocketing? This is one of the biggest challenges Alex Shaw, Member of Technical Staff at The Laude Institute, co-creator of Terminal-Bench, and creator of the Harbor Agentic Evaluation Framework (https://lnkd.in/gzzqRSjj), breaks down with Megan Risdal, Product Lead at Kaggle, and Walter Reade, Technical Lead for Kaggle Competitions for last week’s Kaggle Conversations episode. Alex explains how providing creators with proper tools can help solve the saturation crisis, and also addresses: • The impact of agent scaffolding on performance • How Harbor standardizes data for frontier labs • His own career trajectory and role at the Laude Institute    • The Laude Institute’s multi-year "Moonshot" and "Slingshot" programs Check out this featured clip 👇

Missed the 5-Day AI Agents: Intensive Vibe Coding Course with Google? You can now access the entire course at your own pace with the new Kaggle Learn Guide. Explore the foundations, architecture, and practical development of AI agents through: • Summary podcasts and whitepapers for each day's topic • Hands-on codelabs to apply what you learn • Recorded livestream sessions from every day of the course Start learning here: https://lnkd.in/eWSKJTbV 

Scientists use 3D microscopy to study cells, but turning these massive images into insights represents a significant amount of time during research. 🔬 That’s why Chan Zuckerberg Biohub is partnering with Kaggle to launch the Biohub - Cell Tracking During Development competition. Build machine learning models to identify cell divisions, track changes over time, and reconstruct lineages in real 3D data. Prize Pool: $60,000 Entry Deadline: September 22, 2026 These models will help automate a huge challenge in biological research, making it easier for researchers to study how cells grow, interact, and change. You'll be working with: • Real 3D time-lapse microscopy datasets • Dense cell populations, cell divisions, and challenging biological structures • Methods that can accurately reconstruct cell lineages Join the competition here: https://lnkd.in/eiNfb26x