home.social

#patents — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #patents, aggregated by home.social.

  1. I posted this about a year ago. Might be a good time to repost it.

    "Search With Stateful Chat" patent (Cf. patents.google.com/patent/US20 ) - appears to describe the Gemini app for smartphones.

    "Method for Text Ranking with Pairwise Ranking Prompting" (Cf. patents.google.com/patent/US20 ) - documents an experimental process described in this research paper titled "Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting" (Cf. arxiv.org/pdf/2306.17563 ). There is no indication this was introduced into a live agentic system like Gemini.

    "User Embedding Models for Personalization of Sequence Processing Models" (Cf. patents.google.com/patent/WO20 ) - documents an experimental process for improving recommender (sub-)systems (like movie searches) that incorporate large language models. The process is described in this research paper titled "User Embedding Model for Personalized Language Prompting" (Cf. arxiv.org/pdf/2401.04858 ).

    "Systems and methods for prompt-based query generation for diverse retrieval" (Cf. patents.google.com/patent/WO20 ) - updates a 2022 patent for a process named PROMPTAGATOR that generates queries more efficiently based on a small number of examples, as described in this research paper titled "Promptagator - Few-shot Dense Retrieval from 8 Examples" (Cf. arxiv.org/pdf/2209.11755 ). This could be used to generate query fan-outs (but query fan-out has been used in multiple systems at least since the 1990s, so there are many implementations).

    "Instruction Fine-Tuning Machine-Learned Models Using Intermediate Reasoning Steps" (Cf. patents.google.com/patent/US20 ) - documents an older method for fine-tuning instructions submitted to LLMs, as described in this 2022 research paper titled "Scaling Instruction-Finetuned Language Models" (Cf. jmlr.org/papers/volume25/23-08 ). The work has been superseded by this paper titled "Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models" (Cf. arxiv.org/pdf/2305.14705 ).

    This is the AI Overviews patent, titled "Generative summaries for search results" (Cf. patents.google.com/patent/US11 )

    #google #aioverviews #aimode #machinelearning #search #searchengines #generativesearch #seo #searchengineoptimization #webmarketing #digitalmarketing #ai #patents

    seo-theory.com/how-to-read-pat

  2. I posted this about a year ago. Might be a good time to repost it.

    "Search With Stateful Chat" patent (Cf. patents.google.com/patent/US20 ) - appears to describe the Gemini app for smartphones.

    "Method for Text Ranking with Pairwise Ranking Prompting" (Cf. patents.google.com/patent/US20 ) - documents an experimental process described in this research paper titled "Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting" (Cf. arxiv.org/pdf/2306.17563 ). There is no indication this was introduced into a live agentic system like Gemini.

    "User Embedding Models for Personalization of Sequence Processing Models" (Cf. patents.google.com/patent/WO20 ) - documents an experimental process for improving recommender (sub-)systems (like movie searches) that incorporate large language models. The process is described in this research paper titled "User Embedding Model for Personalized Language Prompting" (Cf. arxiv.org/pdf/2401.04858 ).

    "Systems and methods for prompt-based query generation for diverse retrieval" (Cf. patents.google.com/patent/WO20 ) - updates a 2022 patent for a process named PROMPTAGATOR that generates queries more efficiently based on a small number of examples, as described in this research paper titled "Promptagator - Few-shot Dense Retrieval from 8 Examples" (Cf. arxiv.org/pdf/2209.11755 ). This could be used to generate query fan-outs (but query fan-out has been used in multiple systems at least since the 1990s, so there are many implementations).

    "Instruction Fine-Tuning Machine-Learned Models Using Intermediate Reasoning Steps" (Cf. patents.google.com/patent/US20 ) - documents an older method for fine-tuning instructions submitted to LLMs, as described in this 2022 research paper titled "Scaling Instruction-Finetuned Language Models" (Cf. jmlr.org/papers/volume25/23-08 ). The work has been superseded by this paper titled "Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models" (Cf. arxiv.org/pdf/2305.14705 ).

    This is the AI Overviews patent, titled "Generative summaries for search results" (Cf. patents.google.com/patent/US11 )

    #google #aioverviews #aimode #machinelearning #search #searchengines #generativesearch #seo #searchengineoptimization #webmarketing #digitalmarketing #ai #patents

    seo-theory.com/how-to-read-pat

  3. I posted this about a year ago. Might be a good time to repost it.

    "Search With Stateful Chat" patent (Cf. patents.google.com/patent/US20 ) - appears to describe the Gemini app for smartphones.

    "Method for Text Ranking with Pairwise Ranking Prompting" (Cf. patents.google.com/patent/US20 ) - documents an experimental process described in this research paper titled "Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting" (Cf. arxiv.org/pdf/2306.17563 ). There is no indication this was introduced into a live agentic system like Gemini.

    "User Embedding Models for Personalization of Sequence Processing Models" (Cf. patents.google.com/patent/WO20 ) - documents an experimental process for improving recommender (sub-)systems (like movie searches) that incorporate large language models. The process is described in this research paper titled "User Embedding Model for Personalized Language Prompting" (Cf. arxiv.org/pdf/2401.04858 ).

    "Systems and methods for prompt-based query generation for diverse retrieval" (Cf. patents.google.com/patent/WO20 ) - updates a 2022 patent for a process named PROMPTAGATOR that generates queries more efficiently based on a small number of examples, as described in this research paper titled "Promptagator - Few-shot Dense Retrieval from 8 Examples" (Cf. arxiv.org/pdf/2209.11755 ). This could be used to generate query fan-outs (but query fan-out has been used in multiple systems at least since the 1990s, so there are many implementations).

    "Instruction Fine-Tuning Machine-Learned Models Using Intermediate Reasoning Steps" (Cf. patents.google.com/patent/US20 ) - documents an older method for fine-tuning instructions submitted to LLMs, as described in this 2022 research paper titled "Scaling Instruction-Finetuned Language Models" (Cf. jmlr.org/papers/volume25/23-08 ). The work has been superseded by this paper titled "Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models" (Cf. arxiv.org/pdf/2305.14705 ).

    This is the AI Overviews patent, titled "Generative summaries for search results" (Cf. patents.google.com/patent/US11 )

    #google #aioverviews #aimode #machinelearning #search #searchengines #generativesearch #seo #searchengineoptimization #webmarketing #digitalmarketing #ai #patents

    seo-theory.com/how-to-read-pat

  4. I posted this about a year ago. Might be a good time to repost it.

    "Search With Stateful Chat" patent (Cf. patents.google.com/patent/US20 ) - appears to describe the Gemini app for smartphones.

    "Method for Text Ranking with Pairwise Ranking Prompting" (Cf. patents.google.com/patent/US20 ) - documents an experimental process described in this research paper titled "Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting" (Cf. arxiv.org/pdf/2306.17563 ). There is no indication this was introduced into a live agentic system like Gemini.

    "User Embedding Models for Personalization of Sequence Processing Models" (Cf. patents.google.com/patent/WO20 ) - documents an experimental process for improving recommender (sub-)systems (like movie searches) that incorporate large language models. The process is described in this research paper titled "User Embedding Model for Personalized Language Prompting" (Cf. arxiv.org/pdf/2401.04858 ).

    "Systems and methods for prompt-based query generation for diverse retrieval" (Cf. patents.google.com/patent/WO20 ) - updates a 2022 patent for a process named PROMPTAGATOR that generates queries more efficiently based on a small number of examples, as described in this research paper titled "Promptagator - Few-shot Dense Retrieval from 8 Examples" (Cf. arxiv.org/pdf/2209.11755 ). This could be used to generate query fan-outs (but query fan-out has been used in multiple systems at least since the 1990s, so there are many implementations).

    "Instruction Fine-Tuning Machine-Learned Models Using Intermediate Reasoning Steps" (Cf. patents.google.com/patent/US20 ) - documents an older method for fine-tuning instructions submitted to LLMs, as described in this 2022 research paper titled "Scaling Instruction-Finetuned Language Models" (Cf. jmlr.org/papers/volume25/23-08 ). The work has been superseded by this paper titled "Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models" (Cf. arxiv.org/pdf/2305.14705 ).

    This is the AI Overviews patent, titled "Generative summaries for search results" (Cf. patents.google.com/patent/US11 )

    #google #aioverviews #aimode #machinelearning #search #searchengines #generativesearch #seo #searchengineoptimization #webmarketing #digitalmarketing #ai #patents

    seo-theory.com/how-to-read-pat

  5. I posted this about a year ago. Might be a good time to repost it.

    "Search With Stateful Chat" patent (Cf. patents.google.com/patent/US20 ) - appears to describe the Gemini app for smartphones.

    "Method for Text Ranking with Pairwise Ranking Prompting" (Cf. patents.google.com/patent/US20 ) - documents an experimental process described in this research paper titled "Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting" (Cf. arxiv.org/pdf/2306.17563 ). There is no indication this was introduced into a live agentic system like Gemini.

    "User Embedding Models for Personalization of Sequence Processing Models" (Cf. patents.google.com/patent/WO20 ) - documents an experimental process for improving recommender (sub-)systems (like movie searches) that incorporate large language models. The process is described in this research paper titled "User Embedding Model for Personalized Language Prompting" (Cf. arxiv.org/pdf/2401.04858 ).

    "Systems and methods for prompt-based query generation for diverse retrieval" (Cf. patents.google.com/patent/WO20 ) - updates a 2022 patent for a process named PROMPTAGATOR that generates queries more efficiently based on a small number of examples, as described in this research paper titled "Promptagator - Few-shot Dense Retrieval from 8 Examples" (Cf. arxiv.org/pdf/2209.11755 ). This could be used to generate query fan-outs (but query fan-out has been used in multiple systems at least since the 1990s, so there are many implementations).

    "Instruction Fine-Tuning Machine-Learned Models Using Intermediate Reasoning Steps" (Cf. patents.google.com/patent/US20 ) - documents an older method for fine-tuning instructions submitted to LLMs, as described in this 2022 research paper titled "Scaling Instruction-Finetuned Language Models" (Cf. jmlr.org/papers/volume25/23-08 ). The work has been superseded by this paper titled "Mixture-of-Experts Meets Instruction Tuning: A Winning Combination for Large Language Models" (Cf. arxiv.org/pdf/2305.14705 ).

    This is the AI Overviews patent, titled "Generative summaries for search results" (Cf. patents.google.com/patent/US11 )

    #google #aioverviews #aimode #machinelearning #search #searchengines #generativesearch #seo #searchengineoptimization #webmarketing #digitalmarketing #ai #patents

    seo-theory.com/how-to-read-pat

  6. 🎉 Ultimul brevet pentru MPEG-4 Part 2 a expirat! DivX și Xvid sunt acum complet libere la nivel mondial

    O eră importantă din istoria codecurilor video se încheie oficial: ultimul brevet care proteja standardul MPEG-4 Part 2 (cunoscut pentru formatele populare DivX și Xvid din anii 2000) a expirat.

    💡 Ce înseamnă acest lucru?

    🔓 Fără redevențe sau licențiere:
    Producătorii de hardware, dezvoltatorii de aplicații comerciale, proiectele multimedia și distribuțiile Linux pot include acum suport nativ de codare/decodare MPEG-4 Part 2 fără teama de taxe sau procese legate de brevete.

    🐧 Ideal pentru distribuțiile Linux:
    Distribuțiile cu politici extrem de stricte privind software-ul brevetat vor putea include suport de bază pentru aceste formate fără a mai necesita repozitoare terțe.

    🔍 Clarificare importantă:
    • Expirarea se aplică MPEG-4 Part 2 (fisierele vechi .avi, DivX, Xvid).
    • NU se aplică formatului H.264 (MPEG-4 Part 10) sau altor tehnologii moderne utilizate în fișierele .mp4 actuale, care sunt încă protejate de brevete active.

    Pentru utilizatorul obișnuit (care folosește deja player-e precum VLC sau FFmpeg) nimic nu se schimbă peste noapte, însă este o victorie uriașă pentru distribuția de software liber și arhivarea digitală! 🎬

    #MPEG4 #DivX #Xvid #OpenSource #SoftwareFreedom #Patents #Linux #Multimedia #TechNews #VideoCodecs

  7. 104 year old Eddy Goldfarb has invented over 800 iconic toys- some of my favorites, too. He made Stompers, which were HUGE for me as a kid (I still have several). He also made talking teeth, Kerplunk, just TONS of stuff.

    Cool piece via Moss and Fog (the embedded YouTube video here is a cool watch, also): mossandfog.com/104-year-old-ed

    #toys #ClassicToys #EddyGoldfarb #inventor #inventors #patents #games #Stompers #80s #80sToys #VintageToys #history #design #MossAndFog

  8. Archer Aviation buys Lilium’s €18M patent portfolio, gaining key eVTOL tech like ducted fans and flight controls. This boosts Archer’s edge in urban, regional, and defense electric aircraft, turning Lilium’s innovations into future flying products.

    #ArcherAviation #Lilium #eVTOL #ElectricAircraft #Innovation #Aerospace #Patents #TECHi

    Read Full Article Here : - techi.com/archer-acquires-lili

  9. #News out of #Nintendo's #patent #lawsuit against #PocketPair, alleging the #GameMechanics of #Palworld are too similar to #Pokemon.

    We now know the main #patents, all filed in #Japan earlier this year after PalWorld's release...

    (Reminder that #IAmNotALawyer)

    No. 7493117 involves #collision and #pathfinding, and seems to be mostly related to trainer battles and entering battle by throwing one Pokemon at another. It's not clear to me how this is novel or unique enough for a lawsuit.

    No. 7528390 involves player characters mounting creatures and riding them around. As Pokemon is far from the only, first, or even a game with this mechanic for more than a few years, I don't see how they can claim this.

    No. 7545191 is the #PokeBall thing, which #PalBalls are very similar to, so much I was surprised they did that at the game's release (#Bugsnax at least used a non-spherical trap), but again, I'm not sure how this is unique or innovative enough for a lawsuit.

    #Gaming #GameDev

  10. Hi all! I studied computer science with a focus on RTOS and FP/PL, but I graduated law school earlier this year. I practice at a boutique firm that primarily handles IP cases. I foilboard and I play way too many rhythm games in and out of the arcade.

    I contribute to open source projects where I can, and I write up my experience in my digital garden which I’ve been maintaining for 2 years now. It’s also a good place to find usage tips for projects/tools that you might want to use.

    My passion for tech also includes privacy, and I’m an advocate for minimizing your digital footprint. GenAI is a scam and its purveyors are causing real harm while they sell it as hard as they can.

    Follow me for: #selfhosting #digitalgardening #privacy #lawfedi #section230 #copyright #patents #rustlang #gleam #haskell #RSS #neovim #NixOS #zotero #tmux #alacritty #linux #egpu #qemu #arch #archlinux #GNOME #watches #watchmaking #obsidian #obsidianmd #thunderbird #fpv #mechkeys #mechkeeb #mechanicalkeyboard #matrix #signal #fido2 #passkeys #dancerushstardom

  11. #OnThisDay March 21, 1891: Alb. Jeanneret & Frères of Saint-Imier receives Swiss patent 3364 for their "compteur-chronographe" with the famous J-shaped bridge, establishing the roots of the great Excelsior Park chronograph empire #ExcelsiorPark #SaintImier #Patents

  12. ViceSociety revendique des attaques informatiques à l'encontre de:

    #usa #canada #italy #uk #belgium #ecuador #education #ransomware #vicesociety #líder #teachers #community #campus #databreach #financial #vehículos #services #students #electronics #antennen #employees #diritto #konzepte #tuition #payments #digital #leistungen #shops #kundenservice #products #automatie #web #chemie #elektrische #online #microelectronics #actividades #smart #threats #grades #inteligente #portal #insights #balances #college #siniestros #telekommunikation #elektronica #systems #electrónico #virtual #competitivo #careers #mobilfunk #retails #semiconductors #technologies #brevetti #analysis #digital #mercado #telefonische #fysica #autotechnieken #courses #karriere #sanitair #market #robotics #direktvertrieb #innovative #coberturas #webapi #equipayment #business #póliza #cyberaanval #mercati #mobile #consultas #produkte #competitive #autoconectado #property #application #consumers #laws #firms #facturación #serviceleistungen #productos #patents #seguros #electrónica #riesgos #fiduciaros #automation #wiskunde #tecnología #industriële #piattaforma #asistencias #admissions #soluzioni #belgique #italie #informatique