Blog
Notes on teaching robots new tricks
How the studio works, why tiny robots come first, what the checks catch, and how a policy gets from simulation to a servo.
workflowopen-duck-miniSep 29, 2026 · 6 min readFrom a sentence to a walking duck: how kualia turns a plain-language brief into a training taskA walk through the whole kualia workflow, from a one-sentence brief to an ONNX policy, using the Open Duck Mini and a small ramp.Read the post 
Sep 27, 2026 · 5 minWhy we start with tiny robotsDesk-sized robots train on CPU in minutes to hours, cost little to own and cross the sim-to-real gap without heroics. That is why kualia's shelf starts with them.
Sep 25, 2026 · 6 min13 checks before a single training stepalia writes the training task, but it does not get to grade its own work. Every task passes 13 sandbox checks before training can start. Here is what each one catches.
Sep 23, 2026 · 5 minFrom simulation to servo: running your ONNX policy on a real SO-101What is inside a kualia ONNX export, how to read its metadata, a minimal onnxruntime loop for the SO-101, and the sim-to-real checklist before you power the servos.
Sep 21, 2026 · 5 minLet your coding agent train the robotkualia is a remote MCP server. Point Claude Code, Cursor or Codex at it and your agent can pick a robot, build the task, train it and fetch the ONNX policy from your editor.Research digest
Short reads on new robotics and AI papers, written by our research assistant and reviewed before publishing. 1,060 so far.
Page 47 of 59
Mar 28, 2026Physics as an Interface: Bridging Continuous Actions and Visual Generation in RealWonderCurrent video diffusion models excel at synthesizing visually coherent sequences but operate fundamentally as passive projection engines. They lack causal understanding of how physical interventions reshape scenes, rende Mar 27, 2026 · TR Mar 27, 2026Decoupling Abstraction from Execution: How Diffusion-Based Tokenization Resolves the Efficiency-Fidelity Trade-off in Motion GenerationHuman motion generation sits at a crossroads between two competing methodologies. On one side, continuous diffusion models capture the fluid dynamics and physical plausibility of human movement with remarkable fidelity. Mar 27, 2026 · TRKlinik Çalışma Verileri ile Gözlemsel Veritabanlarını Birleştirmenin Yeni Yolu: Kalibre Hizalama ile Etki Tahmini Giriş Hassas tıp (precision medicine) çağında, tedavilerin bireysel hasta özelliklerine göre uyarlanması kaçınılmaz bir gereksinim haline geldi. Bu uyarlamanın temelini oluşturan heterogeneous treatment effects (HTEs) v Mar 27, 2026Beyond Imputation: Embedding Alignment for Causal Inference Under Covariate MismatchThe integration of randomized controlled trials with observational data represents one of the most pressing methodological challenges in modern biostatistics. While RCTs maintain their status as the definitive source for Mar 27, 2026 · TRÇok Dilli Transformer Modellerinde Gizli Diller Arası Mesafe: Attention Transport Distance ve Ölçüm Sorunu Giriş İnsan dilleri arasındaki benzerlik ve farklılıkları ölçmek, dilbilimin temel sorularından biri olarak tarih öncesi göçlerin izini sürmekten yapay zeka sistemlerinin çok dilli yeteneklerini optimize etmeye kadar ge Mar 27, 2026Attention Transport Distance: A Critical Appraisal of Geometric Measures in Multilingual ModelsThe quantification of linguistic distance has long occupied a central position in historical linguistics and typology. Traditional methods rely on cognate identification, sound correspondences, or curated feature databas Mar 27, 2026 · TRGörüntü Düzenleme Modellerinin Gizli Dünya Anlayışı: Edit Spillover Olgusu Üzerine Bir İnceleme Giriş İnstruction-based görüntü düzenleme modelleri, kullanıcıların doğal dil komutlarıyla görsel içerikleri manipüle etmesine olanak tanıyan, diffusion (dağınım) tabanlı sistemlerin en dikkat çekici uygulamalarından bi Mar 27, 2026Edit Spillover as a Cognitive Probe: Rethinking Errors in Image Editing ModelsWhen an instruction following image editing model changes "Old Wang Noodle House" to "Xinhua Bookstore" on a storefront signboard, and simultaneously alters "Old Wang Special Noodles" to "Xinhua Special Noodles" on a dis Mar 27, 2026 · TRYüz Merkezli Ölçek Bilinçli Koşullandırma: Portrait Video Üretiminde Kamera Kontrolünün Yeniden Tanımlanması Giriş Video üretimi modellerinde kontrollü kamera hareketi, özellikle insan portrelerinin bulunduğu sahnelerde, bilgisayarlı görü ve grafik alanının en zorlu problemlerinden biri olarak öne çıkıyor. Kullanıcıların kayna Mar 27, 2026Scale-Aware Conditioning and the Geometry of Control in Portrait Video GenerationThe generation of portrait videos under arbitrary camera trajectories represents one of the more subtle challenges in contemporary computer vision. While large video generation models have demonstrated remarkable capabil Mar 27, 2026 · TRVideoRAG: Video Kütüphanelerinde Retrieval-Augmented Generation ve Zamansal Süreklilik Problemi Giriş Retrieval-Augmented Generation (RAG), büyük dil modellerinin parametrik bilgisini harici kaynaklarla zenginleştiren güçlü bir strateji olarak öne çıkıyor. Geleneksel olarak metin odaklı işleyen bu yaklaşım, son dö Mar 27, 2026Beyond Keyframes: Temporal Reasoning Challenges in VideoRAG and the Future of Multimodal RetrievalRetrieval Augmented Generation has evolved from textual corpora to multimodal repositories, yet the transition to video presents unique representational challenges. The recent work by Jeong et al., titled "VideoRAG: Retr Mar 27, 2026 · TRLuMamba: Elektrot Topolojisi Bağımsız ve Verimli EEG Modellemesi için Yeni Bir Mimari Giriş: EEG Temel Modellerinin İki Temel Sorunu Elektroensefalografi (EEG) temel modelleri (foundation models), klinik tanıdan beyin-bilgisayar arayüzlerine kadar geniş bir yelpazede uygulama buluyor. Ancak bu alanda çal Mar 27, 2026LuMamba: Rethinking EEG Foundation Models Through Geometric Invariance and State-Space EfficiencyThe proliferation of foundation models in medical signal processing has encountered a fundamental tension in electroencephalography. On one hand, transformer architectures have demonstrated remarkable representational ca Mar 27, 2026 · TRSparkVSR ve Etkileşimli Video Süper Çözünürlük: Seyrek Anahtar Karelerle Kullanıcı Kontrollü Yükseltme Giriş Video süper çözünürlük (Video Super-Resolution, VSR) alanında son yıllarda elde edilen ilerlemelere rağmen, mevcut sistemlerin büyük çoğunluğu hala black box olarak çalışmaktadır. Kullanıcılar düşük çözünürlüklü ( Mar 27, 2026 Mar 27, 2026 · TR