© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    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. 128012
    Prev
    Next
    INDEX
    Explanations

    user engagement and user identifiers in conversations

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

    tokens that mark speakers or direct address—turn/utterance starts and first- or second-person references (speaker/role tokens like "model", "An", and pronouns like "I"/"you").

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
    AndEndTag
    -0.72
     يتيمه
    -0.72
    
    -0.68
     autorytatywna
    -0.61
     saites
    -0.60
    IsContent
    -0.60
     للمعارف
    -0.60
    antwoorde
    -0.53
     surla
    -0.53
    skjaer
    -0.52
    POSITIVE LOGITS
     coleg
    0.35
    UnusedPrivate
    0.32
     remaja
    0.31
    sscanf
    0.31
    ParallelGroup
    0.30
     drama
    0.30
     paranoia
    0.30
     Remind
    0.30
    parseFrom
    0.30
    isNaN
    0.29
    Activations Density 0.002%

    No Known Activations