© 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. 36-GEMMASCOPE-RES-16K
    11. 14920
    Prev
    Next
    INDEX
    Explanations

    expressions of personal admission or acknowledgment, particularly regarding guilt or involvement in wrongdoing

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_36/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.36.hook_resid_post
    Hook Layer
    36
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ymce
    -0.39
    Datuak
    -0.37
     GenerationType
    -0.35
    Див
    -0.35
    ChildScrollView
    -0.34
    看去
    -0.34
    tagext
    -0.34
    JoinTable
    -0.34
     []).
    -0.34
     calendriers
    -0.33
    POSITIVE LOGITS
     admit
    2.91
     admitting
    2.78
     admitted
    2.70
     admits
    2.61
     admission
    2.58
     Admit
    2.53
     admitir
    2.30
     acknowledge
    2.27
     Admission
    2.22
    承认
    2.22
    Activations Density 0.887%

    No Known Activations