X-raying a Transformer Forward Pass
luizspies
• NEW Articles from Team or Enterprise organizations will get promoted to the main section.
Compressing visual tokens in vision-language models: 3x more requests per GPU on Qwen2-VL
datas3nt
• Intel XPU Kernel Skill: LLM-driven Triton kernel optimization for the Hugging Face Kernel Hub
danf
• • 7Don’t Waste Tokens on Data Entry: Tag Customer Reviews Overnight with ZeroGPU Batch API
its-maddy-a
• • 1NanoMathDataset: um dataset open-source de matemática
pedrodev2026
• Party is over: regularizing ColBERT models to fix efficient ANN methods
lightonai
• • 17The Method Behind OUROBOROS: Make the Reward Hard to Fool
YMRohit
• curbcheck: can a small VLM tell you if you can legally park in San Francisco?
shubhamgoel27
• Signal Garden: A Game Engine That Keeps Mutating
build-small-hackathon
• • 2Noteworthy
build-small-hackathon
• Yui Home Assistant — teaching a 3B model to write Home Assistant automations that actually *work*
build-small-hackathon
• Reading Buddy: The Quiet Companion That Reads Alongside You
jrpaul08
• Household Food Waste Forensics: Smart Tracking for Sustainable Kitchens
abhishekkataria16
• My First MO§ES™ SigRank — the diagnostic x-ray of the token economy
build-small-hackathon
• Keep the model on one axis: building *Penny Stock Persuasion* on a small Nemotron
Han-Solo
• • 1Moku: The First Lie in the Canopy
build-small-hackathon
• Memory Quilt: Weaving Family Stories into Digital Art
abhishekkataria16
• Elder Care Copilot: Bringing Order to the Chaos of Caregiving Paperwork
build-small-hackathon
• I built a tool that tells you exactly why you'll be rejected — before you apply
build-small-hackathon
• • 1Field Notes: building a shop-floor AI on a small local model
build-small-hackathon
• • 1