HN Explorer

DISCURSIVE

↗ HN profile [ 55.1 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. Nvidia's Openness Promise Leaves Out Hosting, Gating and Licensing (topicqueue.substack.com)
    1 points 1
  2. Perceptron Ships 36B Robot Weights Under Apache-2.0 (topicqueue.substack.com)
    2 points 0
  3. Model Success Rate May Count the Wrong Runs (topicqueue.substack.com)
    1 points 0
  4. RoboDojo Scores 32 of 34 Robot Policies Under 10% Success (topicqueue.substack.com)
    1 points 0
  5. Dyna-2: A Million Hours of Human Video, Zero Robot Frames (topicqueue.substack.com)
    1 points 0
  6. Gemini Robotics Released One Policy from Feet to Fingertips (topicqueue.substack.com)
    2 points 0
  7. 5k Hours of Tactile Data, and Four Ways to Get It (topicqueue.substack.com)
    1 points 0
  8. Nvidia Puts a 4B World Model on the Robot [Weekly Physical AI Roundup] (topicqueue.substack.com)
    2 points 0
  9. Turning robot video into training-ready data (eventual.ai)
    2 points 0
  10. Xiaomi Opens a 38B World Model Built to Generate Robot Data (topicqueue.substack.com)
    4 points 1
  11. Querying Physical AI Data with Daft (eventual.ai)
    1 points 0
  12. LeRobot v0.6.0 Adds World Models, Reward Models, and an Open GR00T (topicqueue.substack.com)
    1 points 0
  13. A Push on Tactile Data, and a Warning That the Benchmark Was Leaking (topicqueue.substack.com)
    2 points 0
  14. Hours of Humanoid Teleop, Recorded in Real Homes (topicqueue.substack.com)
    3 points 0
  15. Knowledge curation (not search) is the AI big data problem (daft.ai)
    3 points 0
  16. Cutting LLM Batch Inference Time by Half with Dynamic Prefix Bucketing (daft.ai)
    2 points 0
  17. Agentic systems are just query engines for unstructured data (daft.ai)
    3 points 0
  18. Benchmark: Spark vs. Ray Data vs. Daft on Multimodal Workloads (daft.ai)
    1 points 0
  19. Show HN: Flotilla, the Distributed Engine for Multimodal Pipelines (daft.ai)
    2 points 0
  20. Profiling multimodal workloads: lessons from Daft (daft.ai)
    1 points 0
  21. Processing 24T tokens for LLM training with 0 crashes (what made it possible) (daft.ai)
    1 points 0
  22. We Hit 100% GPU Utilization–and Then Made It 3× Faster by Not Using It (daft.ai)
    17 points 5
  23. Elasticsearch Was Great, but Vector Databases Are the Future (thenewstack.io)
    16 points 7
  24. BGE-M3 vs. Splade (zilliz.com)
    1 points 0
  25. Enhance information retrieval with learned sparse embeddings (zilliz.com)
    1 points 0