[{"data":1,"prerenderedAt":119},["ShallowReactive",2],{"project-41905357":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":9,"maker":10,"posted_at":11,"year":12,"points":13,"comments":14,"rank_in_year":15,"shown_in_year":16,"top":17,"tier":18,"category":19,"subcategory":20,"format":21,"keywords":22,"reception":27,"sentiment":9,"stance_mix":28,"scored_comments":29,"aspects":31,"strip":32,"post_text":34,"what":35,"said":54,"maker_in_thread":68,"shape":73,"grid":76,"attempts":81,"same_week":82,"maker_shows":33,"maker_hidden":111,"reader_version":112,"card":113},"41905357","Forest Friends","A book on LLM system evals with a bear and fox","title","Show HN: We wrote a book on LLM system evals with a bear and fox",null,"iamwil",1729525361,2024,11,2,2489,3515,71,"shown","learning_science","technical zine on llm evals","other",[23,24,25,26],"llm evals","system evals","llm evaluation","ai engineering","quiet",{"positive":29,"neutral":30,"mixed":30,"critical":30},1,0,{},[33,30],5,"Hey all!\n\n@sridatta and I wrote a book\u002Fzine called Forest Friends on system evals for LLM-driven apps. But it's a bit more whimsical, a bit more visual, and very much inspired by the meme of LLMs being a shoggoth polished into a smiley face with RLHF.\n\nLLM system evals are important as companies move past the flashy AI demos to reliable production apps. System evals keep coming up as the answer for what you \"should do\", but it's not exactly a standard part of the software engineering toolkit.\n\nSo we pulled from @sridatta's seven years as a research engineer at Google, plus a ton of best practi",{"problem":36,"audience":42,"pricing":9,"motivation":46,"time_to_build":9,"stack":50,"openness":51,"stage":52,"team_size":53},{"value":37,"source":38,"evidence":39},"Teaching engineers how to write system evals for LLM-driven apps","maker",[40],{"id":41},"post",{"value":43,"source":38,"evidence":44},"Engineers building LLM-driven production apps",[45],{"id":41},{"value":47,"source":38,"evidence":48},"Make eval practices approachable and turn deep expertise into an entrepreneurial project",[49],{"id":41},[],"unknown","launched","team",{"praise":55,"objections":64,"requests":65,"questions":66,"alternatives":67},[56],{"value":57,"source":58,"evidence":59},"Readers find it fun, clear and useful for getting the basics right","commenter",[60],{"id":61,"by":62,"text":63},"41906114","jcgr","amazing, thanks for the pdf.. just starting to do my evals",[],[],[],[],{"comments":29,"answered":30,"key":69},[70],{"id":71,"text":72},"41906956","Thanks! Let me know if there's anything I can improve on it, or anything that you particularly liked\u002Fdisliked about the content and reading experience.",{"commenters":14,"max_depth":14,"maker_replies":29,"first_comment_mins":74,"lifespan_hours":75},84,3,[77,79],[61,30,33,78,62,30],"project\u002Fpositive\u002Fnone",[71,30,30,80,10,29],"commenter\u002Fneutral\u002Fnone",[],[83,88,93,97,102,106],{"id":84,"name":85,"points":86,"category":87},"41893524","HN Update",644,"ai",{"id":89,"name":90,"points":91,"category":92},"41914544","Rust Web Framework",488,"languages",{"id":94,"name":95,"points":96,"category":87},"41929804","Pablonet",472,{"id":98,"name":99,"points":100,"category":101},"41915958","Counter-Strike Football",420,"games",{"id":103,"name":104,"points":105,"category":87},"41926770","Agent.exe",406,{"id":107,"name":108,"points":109,"category":110},"41903864","Epublifier",290,"productivity",false,"v2",{"id":114,"name":5,"line":6,"ai":111,"cat":19,"year":12,"points":13,"comments":14,"rec":27,"s":9,"top":17,"kw":115,"maker":10,"domain":116,"tier":18,"attempts":30,"strip":117,"url":-1,"mix":118,"praise":57,"objection":-1,"sub":20},41905357,[23,24,25,26],"","50",[29,30,30,30],1793991103178]