© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 30-GEMMASCOPE-RES-16K
    11. 12225
    Prev
    Next
    INDEX
    Explanations

    instances of personal attacks and derogatory language

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

    Negative or critical statements

    np_acts-logits-general · gemini-2.0-flash

    attacks, insults, and criticism

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 30-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_30/width_16k/average_l0_120
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.30.hook_resid_post
    Hook Layer
    30
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    醐
    -0.46
     houſe
    -0.44
    nodoc
    -0.44
    fele
    -0.43
     Admit
    -0.43
     confess
    -0.43
    RenderAtEndOf
    -0.41
     AppBar
    -0.41
     experimento
    -0.40
     admit
    -0.40
    POSITIVE LOGITS
     insults
    1.16
     criticism
    1.06
     insulting
    1.05
     criticisms
    1.05
     criticizing
    0.99
     slander
    0.99
     insult
    0.98
     derogatory
    0.97
     criticize
    0.95
    Criticism
    0.94
    Activations Density 0.738%

    No Known Activations