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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 92138
    Prev
    Next
    INDEX
    Explanations

    instances of variables, especially related to creating or managing new objects

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

    state or status keywords

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     otherwise
    -0.35
     either
    -0.32
     also
    -0.32
     if
    -0.31
     with
    -0.31
     as
    -0.31
    ,
    -0.31
     for
    -0.29
     are
    -0.28
     have
    -0.28
    POSITIVE LOGITS
    ſſung
    0.84
    pingente
    0.82
     wikipagina
    0.81
    majánló
    0.81
    ðsíða
    0.81
     Verſ
    0.81
    pecabe
    0.80
    <unused74>
    0.79
    <unused68>
    0.79
    <unused16>
    0.79
    Activations Density 0.013%

    No Known Activations