© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 121305
    Prev
    Next
    INDEX
    Explanations

    phrases and terms related to assigning blame and responsibility

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

    assigning blame or fault

    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
    دانشنامهٔ
    -0.53
    andoah
    -0.49
     Meksiku
    -0.49
    tschaft
    -0.43
    Binding
    -0.42
    ivamente
    -0.42
    rayage
    -0.41
    AnimationsModule
    -0.41
    Pencil
    -0.40
    icherheit
    -0.40
    POSITIVE LOGITS
     blame
    0.93
     blaming
    0.77
    blame
    0.75
     blamed
    0.74
     blames
    0.69
    Blame
    0.64
     Blame
    0.62
     culpa
    0.54
     fault
    0.53
     responsabilité
    0.53
    Activations Density 0.066%

    No Known Activations