[{"data":1,"prerenderedAt":531},["ShallowReactive",2],{"project-42507116":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":29,"sentiment":30,"stance_mix":31,"scored_comments":36,"aspects":37,"strip":44,"post_text":45,"what":46,"said":84,"maker_in_thread":216,"shape":229,"grid":234,"attempts":493,"same_week":494,"maker_shows":39,"maker_hidden":523,"reader_version":524,"card":525},"42507116","PaperMatch","A website to semantically search ArXiv papers","title","Show HN: I made a website to semantically search ArXiv papers","https:\u002F\u002Fpapermatch.mitanshu.tech\u002F","papermatch.mitanshu.tech","Quizzical4230",1735105452,2024,324,104,175,3515,5,"legend","languages","semantic paper search","web_app",[24,25,26,27,28],"semantic search","arxiv","literature review","embeddings","vector search","mixed",0.129,{"positive":32,"neutral":33,"mixed":34,"critical":35},16,18,2,9,45,{"idea":38,"execution":40,"performance":41},{"positive":35,"critical":39},1,{"positive":39,"critical":18},{"positive":42,"critical":43},0,3,[18,18,43,39,42,43,18,18,42,42,43,43,43,42,43,43,43,39,43,39,18,18,39,39,18,18,39,18,39,42,42,18,43,43,43,18,18,18,43,39],"As a grad student (and an ADHDer), I had trouble doing literature review systematically. To combat this, I made a website that finds similar papers using the meaning of the thing I am looking for.\n\nI used MixedBread's [^1] embedding model to generate vectors from the abstracts. I store and search similar vectors using Milvus [^2] and finally use Gradio [^3] to serve the frontend. I update the vector database weekly by pulling the metadata dataset from Kaggle [^4].\n\nTo speed up the search process on my free oracle instance, I binarise the embeddings and use Hamming distance as a metric.\n\nI woul",{"problem":47,"audience":53,"pricing":57,"motivation":58,"time_to_build":57,"stack":62,"openness":81,"stage":82,"team_size":83},{"value":48,"source":49,"evidence":50},"Finding similar academic papers by meaning during literature review","maker",[51],{"id":52},"post",{"value":54,"source":49,"evidence":55},"Grad students and researchers doing literature reviews",[56],{"id":52},null,{"value":59,"source":49,"evidence":60},"Maker struggled with systematic literature review as a grad student with ADHD",[61],{"id":52},[63,67,71,75],{"value":64,"source":49,"evidence":65},"Milvus vector database",[66],{"id":52},{"value":68,"source":49,"evidence":69},"Gradio frontend",[70],{"id":52},{"value":72,"source":49,"evidence":73},"MixedBread mxbai-embed-large-v1 embedding model",[74],{"id":52},{"value":76,"source":49,"evidence":77},"Hugging Face Spaces for weekly update",[78],{"id":79,"by":11,"text":80},"42513241","1. The model size was small enough to process the corpus fast-ish using the limited resources I have. They also support MRL and binary embeddings which help would be helpful in…","open_source","launched","solo",{"praise":85,"objections":104,"requests":139,"questions":171,"alternatives":190},[86,97],{"value":87,"source":88,"evidence":89},"Semantic results are useful for discovering related papers","commenter",[90,94],{"id":91,"by":92,"text":93},"42515761","curious_cat_163","This is great! I just tried some queries and the results were pretty decent, in terms of semantics. But, just thinking of it as a user, if this were to be part of my daily…",{"id":95,"by":11,"text":96},"42509343","All benefits depend on the ability of the embedding model. Semantic embeddings understand nuances, so they can match abstracts that align conceptually even if no exact keywords…",{"value":98,"source":88,"evidence":99},"Useful companion to bioRxiv and medRxiv versions",[100],{"id":101,"by":102,"text":103},"42509764","madbutcode","This looks great! I have used the biorXiv version of papermatch and it gives pretty good results!",[105,116,123,129],{"value":106,"source":88,"evidence":107},"Searches for specific terms like leaky relu return irrelevant results",[108,112],{"id":109,"by":110,"text":111},"42516777","kouteiheika","Feedback: first thing I tried is searching for \"leaky relu\" and I got a bunch of results related to fluids, which is... not very relevant. (: Compare that to scholar which…",{"id":113,"by":114,"text":115},"42517240","zzyzek","This seems like a cool idea, thanks for creating it! Some feedback: I tried searching for \"wave function collapse algorithm\", \"gumin wave function collapse\", \"wfc\" and \"model…",{"value":117,"source":88,"evidence":118},"Author search does not work",[119],{"id":120,"by":121,"text":122},"42513692","serial_dev","I tried a simple search by author and it didn’t work. All the fancy stuff is great, but I’d expect the basics still work, in the end it’s a search engine for papers.",{"value":124,"source":88,"evidence":125},"Cloudflare challenge screen is a barrier",[126],{"id":127,"by":11,"text":128},"42517010","> The Cloudflare challenge screen at the beginning is a dealbreaker. I understand your concern, however, I do not have the know-how to properly combat bots that keep spamming the…",{"value":130,"source":88,"evidence":131},"Binary embeddings reduce recall beyond top 10 results",[132,136],{"id":133,"by":134,"text":135},"42509688","binarymax","That looks great for speed, but what about recall?",{"id":137,"by":11,"text":138},"42509769","That's has a major downgrade. For binary embeddings, the top 10 results are same as fp32, albeit shuffled. However after the 10th result, I think quality degrades quite a bit. I…",[140,151,157,164],{"value":141,"source":88,"evidence":142},"Date or recency filter",[143,147],{"id":144,"by":145,"text":146},"42516810","ProofHouse","I couuld and really use this, but it didn't work for me. And HAS to have a date filter. That is a must maybe with some time based pre-option defaults like HackerNews. Good luck,…",{"id":148,"by":149,"text":150},"42510655","swyx","1. why mixbread's model? 2. how much efficiency gain did you see binarising embeddings\u002Fusing hamming distance? 3. why milvus over other vector stores? 4. did you automate the…",{"value":152,"source":88,"evidence":153},"Keyword and quoted-phrase search",[154],{"id":155,"by":11,"text":156},"42513277","Exactly. Full text paradigm has it's own pros and I believe we need those tools in the new vector search to take full advantage. I am planning to add keywords feature where if a…",{"value":158,"source":88,"evidence":159},"Similar papers link on each result",[160],{"id":161,"by":162,"text":163},"42510370","Maro","Very cool! Add a \"similar papers\" link to each paper, that will make this the obvious way to discover topics by clicking along the similar papers.",{"value":165,"source":88,"evidence":166},"Fix LaTeX rendering in abstracts",[167],{"id":168,"by":169,"text":170},"42512613","venice_benice","interesting project; I’m not really sure how useful it is for field-specific stuff—I'm searching for “image reduction astronomy”, and it shows all sorts of related but not…",[172,183],{"value":173,"source":88,"evidence":174},"Why binarised embeddings and Hamming distance",[175,179],{"id":176,"by":177,"text":178},"42508823","antman","Nice work. Any other technical comments, why did you use those embeddings, did you binarzue them, did you use any dpecial prompts?",{"id":180,"by":181,"text":182},"42509580","andai","Did you notice a difference in performance after binarization? Do you have a way to measure performance?",{"value":184,"source":88,"evidence":185},"How search quality is evaluated",[186],{"id":187,"by":188,"text":189},"42527627","higty","It sounds nice. How do you evaluate the performance of your way against usual embedding?",[191,198,202,209],{"value":192,"source":88,"evidence":193},"Semantic Scholar",[194],{"id":195,"by":196,"text":197},"42507973","shigeru94","Is this similar to https:\u002F\u002Fwww.semanticscholar.org (from Allen Institute for AI) ?",{"value":199,"source":88,"evidence":200},"Google Scholar",[201],{"id":109,"by":110,"text":111},{"value":203,"source":88,"evidence":204},"Research Rabbit",[205],{"id":206,"by":207,"text":208},"42514760","zerop","This looks great, thanks for building this. Something on similar lines which many may link, Research Rabbit - https:\u002F\u002Fwww.researchrabbit.ai\u002F",{"value":210,"source":88,"evidence":211},"Elicit, Undermind, scite",[212],{"id":213,"by":214,"text":215},"42508243","shishy","I enjoy seeing projects like this! If you expand beyond arxiv, keep in mind since coverage matters for lit reviews, unfortunately the big publishers (Elsevier and Springer) are…",{"comments":217,"answered":218,"key":219},40,15,[220,223,226],{"id":221,"text":222},"42509090","Thank you for the appreciation and great feedback! | If you expand beyond arxiv, keep in mind since coverage matters for lit reviews, I do have PaperMatchBio [^1] for bioRxiv and…",{"id":224,"text":225},"42509360","This is quite unique. I believe a custom solution might help you better than Google Scholar.",{"id":227,"text":228},"42513255","So true ;-; I was following the Gmail protocol. I will use > from now on. Happy Holidays :D",{"commenters":230,"max_depth":231,"maker_replies":217,"first_comment_mins":232,"lifespan_hours":233},50,6,254,119,[235,237,239,242,243,245,249,252,253,256,260,261,263,265,266,267,270,271,274,277,281,283,287,289,291,293,296,298,302,303,306,310,311,313,315,319,322,325,328,330,333,335,336,338,341,343,346,348,352,356,358,360,362,365,367,369,373,375,378,381,382,384,386,389,392,394,395,398,402,404,407,409,411,413,416,421,423,425,427,429,431,434,436,438,440,442,443,444,447,449,452,455,459,463,465,469,471,475,477,481,483,486,489,491],[213,42,18,236,214,42],"project\u002Fpositive\u002Fidea",[221,42,42,238,11,39],"maker\u002Fneutral\u002Fnone",[240,42,18,236,241,42],"42509169","eric-burel",[224,42,42,238,11,39],[244,42,18,236,241,42],"42511027",[246,42,42,247,248,42],"42512543","other\u002Fneutral\u002Fnone","dbmikus",[250,42,43,238,251,42],"42512204","mattigames",[227,42,42,238,11,39],[254,42,42,247,255,42],"42515892","immibis",[257,42,39,258,259,42],"42516101","project\u002Fcritical\u002Fexecution","zackmorris",[127,42,42,238,11,39],[262,42,42,247,259,42],"42525292",[264,42,42,247,214,42],"42517340",[148,39,43,238,149,42],[79,39,42,238,11,39],[268,39,18,269,149,42],"42513479","maker\u002Fpositive\u002Fnone",[91,34,18,236,92,42],[272,34,42,273,11,39],"42516117","maker\u002Fneutral\u002Fidea",[275,34,18,236,276,42],"42516619","odyssey7",[278,43,43,279,280,42],"42511558","project\u002Fneutral\u002Fperformance","fasa99",[282,43,42,238,11,39],"42513256",[284,285,42,247,286,42],"42508883",4,"dmezzetti",[288,285,42,247,214,42],"42508934",[290,285,42,247,286,42],"42508968",[292,285,42,238,11,39],"42509216",[294,18,42,247,295,42],"42513591","underlines",[297,18,42,238,11,39],"42517013",[299,231,43,300,301,42],"42509241","project\u002Fneutral\u002Fidea","omarhaneef",[95,231,42,273,11,39],[304,231,43,300,305,42],"42510638","feznyng",[307,231,42,308,309,42],"42511349","commenter\u002Fneutral\u002Fidea","woodson",[155,231,42,273,11,39],[312,231,43,300,305,42],"42515112",[314,231,42,238,11,39],"42515321",[316,317,42,247,318,42],"42508939",7,"namanyayg",[320,317,42,247,321,42],"42509006","bubaumba",[323,317,42,247,324,42],"42509791","imadethis",[326,317,42,247,327,42],"42511023","jondwillis",[195,329,43,300,196,42],8,[331,329,43,300,332,42],"42508222","triilman",[334,329,42,273,11,39],"42509019",[113,35,39,258,114,42],[337,35,42,238,11,39],"42519534",[339,35,18,340,114,42],"42524350","project\u002Fpositive\u002Fexecution",[342,35,42,238,11,39],"42528810",[109,344,39,345,110,42],10,"project\u002Fcritical\u002Fperformance",[347,344,42,238,11,39],"42516926",[349,350,43,300,351,42],"42508125",11,"lgas",[353,350,42,354,355,42],"42508721","commenter\u002Fneutral\u002Fnone","cluckindan",[357,350,42,354,286,42],"42508855",[359,350,42,238,11,39],"42509247",[120,361,39,258,121,42],12,[363,361,42,354,364,42],"42513956","wodenokoto",[161,366,18,236,162,42],13,[368,366,42,238,11,39],"42513258",[370,371,18,236,372,42],"42509442",14,"mskar",[374,371,42,238,11,39],"42509466",[376,371,39,345,377,42],"42515413","OutOfHere",[379,371,42,380,11,39],"42521653","maker\u002Fneutral\u002Fperformance",[187,218,43,279,188,42],[383,218,42,380,11,39],"42546534",[385,32,39,345,377,42],"42515568",[387,32,42,388,11,39],"42517071","project\u002Fmixed\u002Fperformance",[101,390,18,391,102,42],17,"project\u002Fpositive\u002Fnone",[393,390,42,269,11,39],"42512904",[206,33,18,391,207,42],[396,33,42,397,11,39],"42517046","project\u002Fneutral\u002Fnone",[399,400,39,258,401,42],"42508801",19,"mrjay42",[403,400,42,238,11,39],"42509048",[405,406,43,300,321,42],"42508848",20,[408,406,42,273,11,39],"42509147",[410,406,18,236,321,42],"42509465",[412,406,42,300,11,39],"42509500",[414,406,43,397,415,42],"42512184","ttpphd",[417,418,39,419,420,42],"42509211",21,"project\u002Fcritical\u002Fidea","tokai",[422,418,42,354,11,39],"42509225",[424,418,42,354,286,42],"42509253",[426,418,42,354,11,39],"42509372",[428,418,42,354,286,42],"42509402",[430,418,42,354,377,42],"42515427",[432,418,18,236,433,42],"42514759","WolfOliver",[176,435,43,279,177,42],22,[437,435,42,279,11,39],"42509186",[180,439,43,279,181,42],23,[441,439,42,279,11,39],"42509636",[133,439,43,388,134,42],[137,439,42,345,11,39],[445,439,43,279,446,42],"42510224","intalentive",[448,439,42,380,11,39],"42513286",[450,439,43,279,451,42],"42514042","amitness",[453,439,42,454,11,39],"42514993","project\u002Fpositive\u002Fperformance",[456,457,18,391,458,42],"42511281",24,"maCDzP",[460,461,43,397,462,42],"42509275",25,"gaborme",[464,461,42,397,11,39],"42509311",[466,467,18,391,468,42],"42534998",26,"cryptonector",[470,467,42,269,11,39],"42541212",[472,473,18,391,474,42],"42512556",27,"amelius",[476,473,42,247,11,39],"42513261",[478,479,43,397,480,42],"42509123",28,"ukuina",[482,479,42,300,11,39],"42509160",[168,484,43,485,169,42],29,"project\u002Fmixed\u002Fexecution",[487,484,42,488,11,39],"42513315","project\u002Fneutral\u002Fexecution",[144,490,39,258,145,42],30,[492,490,42,397,11,39],"42517022",[],[495,500,505,510,515,519],{"id":496,"name":497,"points":498,"category":499},"42506569","FixBrowser",320,"dev_tools",{"id":501,"name":502,"points":503,"category":504},"42471913","Artemis",312,"productivity",{"id":506,"name":507,"points":508,"category":509},"42497173","isle",251,"games",{"id":511,"name":512,"points":513,"category":514},"42493832","Super Snowflake Maker",228,"creative",{"id":516,"name":517,"points":518,"category":499},"42464310","Mizu.js",225,{"id":520,"name":521,"points":522,"category":509},"42480624","Eonfall",222,false,"v2",{"id":526,"name":5,"line":6,"ai":523,"cat":20,"year":13,"points":14,"comments":15,"rec":29,"s":527,"top":18,"kw":528,"maker":11,"domain":10,"tier":19,"attempts":42,"strip":529,"url":9,"mix":530,"praise":87,"objection":106,"sub":21},42507116,0.13,[24,25,26,27,28],"5531035500333033313155115515100533355531",[32,33,34,35],1794036666299]