© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 49282
    Prev
    Next
    INDEX
    Explanations

    actions and emotions related to pressure and manipulation

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

    succumb to pressure or be easily manipulated

    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
    ValueStyle
    -0.63
     يتيمه
    -0.49
    Bibliograf
    -0.41
    homonymie
    -0.40
     noqa
    -0.40
     purpoſe
    -0.40
    afficheront
    -0.39
     lenker
    -0.39
    verwijspagina
    -0.39
     Exacts
    -0.38
    POSITIVE LOGITS
    SequentialGroup
    0.60
     easily
    0.58
    easily
    0.57
     InputDecoration
    0.56
     weak
    0.51
     fácilmente
    0.51
     cowardice
    0.50
     легко
    0.50
     succumb
    0.50
    Easily
    0.49
    Activations Density 0.091%

    No Known Activations