Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 1234
    Prev
    Next
    INDEX
    Explanations

    content related to gaming performance and technical specifications

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Game modifications or expansions

    np_acts-logits-general · gemini-2.0-flash

    game mods and glitches

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     adopción
    -0.42
     flexível
    -0.42
     flexibilidad
    -0.40
     escalera
    -0.39
     spørs
    -0.39
    enumii
    -0.39
     celo
    -0.39
    æde
    -0.39
     habitudes
    -0.39
     Autónoma
    -0.38
    POSITIVE LOGITS
     modding
    0.72
     mods
    0.71
     modded
    0.65
     gameplay
    0.64
     Gameplay
    0.63
    MLLoader
    0.62
     Mods
    0.59
    minecraftforge
    0.57
     mod
    0.56
     jsPsych
    0.55
    Activations Density 0.179%

    No Known Activations