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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 56398
    Prev
    Next
    INDEX
    Explanations

    references to additional resources or documentation

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

    see or see also

    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
    ########.
    -0.68
    IntoConstraints
    -0.63
    transQ
    -0.60
    cupertino
    -0.59
    stylers
    -0.59
    uxxxx
    -0.58
     continúas
    -0.57
     المعيارى
    -0.57
    RegressionTest
    -0.56
    liceerd
    -0.56
    POSITIVE LOGITS
    see
    1.05
    See
    0.76
     see
    0.74
    SEE
    0.72
     See
    0.68
    siehe
    0.68
     SEE
    0.61
     siehe
    0.58
    Siehe
    0.54
    см
    0.54
    Activations Density 0.005%

    No Known Activations