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. 1229
    Prev
    Next
    INDEX
    Explanations

    concepts related to abstraction and the invisible or unseen forces at play in various contexts

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

    Dark Matter or invisible things

    np_acts-logits-general · gemini-2.0-flash

    intangible, invisible, unseen, imaginary

    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
     Comprometido
    -0.50
    dyž
    -0.47
    exitRule
    -0.45
    andoah
    -0.43
    bitField
    -0.42
    ambién
    -0.42
    pecabezas
    -0.42
     satisfecho
    -0.42
     tertutup
    -0.41
    ſelves
    -0.41
    POSITIVE LOGITS
    VIAF
    0.61
    Autoritní
    0.60
     intangible
    0.60
     invisible
    0.57
     virtual
    0.54
    adaptiveStyles
    0.50
     ethereal
    0.50
     imaginary
    0.48
    invisible
    0.48
     ghosts
    0.47
    Activations Density 0.597%

    No Known Activations