HN Explorer

galsapir

↗ HN profile [ 87.8 ms ] full profile
Karma
0
Created
()
Submissions
0
  1. Giving a domain a hill to climb: benchmarking as data activation (sparsethought.com)
    2 points 7
  2. A bitter lesson for medicine, or a benchmark problem? (sparsethought.com)
    2 points 0
  3. Can LLMs Beat Classical Hyperparameter Optimization Algorithms? (arxiv.org)
    120 points 11
  4. Gemma 4 E4B as a primary local LLM (replaced Qwen) (digg.com)
    2 points 0
  5. PEEK: Give Your Agent an Orientation Cache (MIT CSAIL, Khattab group) (zhuohangu.github.io)
    3 points 0
  6. Hyperagents (Meta Research) (arxiv.org)
    2 points 0
  7. The Unreasonable Effectiveness of HTML (claude.com)
    3 points 2
  8. 1 points 1
  9. The Comparator in Clinical AI (sparsethought.com)
    2 points 1
  10. Borges' cartographers and the tacit skill of reading LM output (galsapir.github.io)
    40 points 5
  11. 1 points 0
  12. Best read of 2026 so far was written in 1880 (galsapir.github.io)
    1 points 1
  13. Anthropic launched community ambassador program (claude.com)
    1 points 1
  14. LLMs as nudging research towards luke-warm middle (nature.com)
    1 points 1
  15. 1 points 0
  16. 1 points 1
  17. 1 points 1
  18. 1 points 1
  19. 1 points 1
  20. How do you evaluate a foundation model before you know what it's for? (galsapir.github.io)
    1 points 1
  21. Ask HN: Anyone using Claude Agent SDK in production?

    We're evaluating agent frameworks for a health AI product and leaning toward Anthropic's Claude Agent SDK. Did a quick POC and liked the simplicity: clean @tool decorator, native MCP support, flat mental model. But I'm…

    1 points 1
  22. Data Activation Thoughts (galsapir.github.io)

    i've been working with healthcare/biobank data and keep thinking about what "data moats" mean now that llms can ingest anything. some a16z piece from 2019 said moats were eroding — now the question seems to be whether…

    21 points 4
  23. 1 points 0