mdda.net - personal website
ResearchTechTagsAbout
Home
Research
Tech
Tags
About

Latest

Blog{mdda} : thought process / brain dump

  • Published on
    December 16, 2025

    Sizing gemma3 for TPUv5-n

    GCPAISprintTPUTPUv5-1TPUv5-8JAXnnxtunixkagglegemma3
    Read more →
  • Published on
    December 15, 2025

    Practical TPU VM on GCP

    GCPAISprintTPUTPUv5-1TPUv5-8JAXnnxtunixkaggle
    Read more →
  • Published on
    December 6, 2025

    Reinforcement Learning for Long-Horizon Multi-Turn Search Agents

    PaperLLMWorkshopRAGAgentsReinforcementLearningRL
    Read more →
  • Published on
    November 26, 2025

    Evolving Reasoning

    LLMSAM3SAM3D-ObjectSAM3D-BodyReasoningEvolutionReinforcement-LearningGRPODistillationOn-Policy
    Read more →
  • Published on
    November 22, 2025

    DevFest SG Talk: Lessons from Reinforcement Learning

    LLMReinforcement-LearningPromptingFineTuningSFTDevFest
    Read more →
  • Published on
    October 15, 2025

    Neural Assets and World Models

    TransformersStableDiffusionDallENanoBananaVOYAGERMinecraftTinyWorldsGENIE
    Read more →
  • Published on
    September 25, 2025

    IPhO Gold using Agentic Gemini

    LLMTransformersAgentsIPhOPhysicsPhysicsOlympiadGemini
    Read more →
  • Published on
    August 27, 2025

    Agent Efficiency, Memory and Confidence

    LLMTransformersAgentsEfficiencyMemoryARPODeepThink
    Read more →
  • Published on
    July 16, 2025

    A Reasoning-Based Approach to Cryptic Crossword Clue Solving

    PaperLLMReasoningFormalisationCryptic-Crosswords
    Read more →
  • Published on
    July 3, 2025

    GPU Kernel Scientist : An LLM-Driven Framework for Iterative Kernel Optimization

    PaperLLMGPUAMDCUDAGeminiWorkshop
    Read more →
All Posts →
mailgithubyoutubelinkedintwitter
Martin Andrews
•
© 2025
•
mdda
Tailwind Nextjs Theme