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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 102320
    Prev
    Next
    INDEX
    Explanations

    terms related to "pedigree" and "pedophilia" as well as any variations of the word "pedal."

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

    ped followed by 'igree' or 'ophile' or 'opening'

    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
    ||}
    -0.59
     للاسماء
    -0.42
     Wirth
    -0.40
    Hii
    -0.40
     Fraser
    -0.39
    Ww
    -0.38
     Connell
    -0.38
    Buat
    -0.38
    ilaian
    -0.37
    OrWhiteSpace
    -0.36
    POSITIVE LOGITS
     Ped
    1.07
     PED
    1.02
    Ped
    1.02
     ped
    0.96
     pedal
    0.93
     Pedal
    0.93
     pedals
    0.90
     Pedagog
    0.84
     pedagogy
    0.82
    PED
    0.81
    Activations Density 0.004%

    No Known Activations