© 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. 2-GEMMASCOPE-RES-16K
    11. 13183
    Prev
    Next
    INDEX
    Explanations

    language reflecting strong negative emotions, particularly hate, as well as references to specific segments or categories

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 2-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_2/width_16k/average_l0_67
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    seen
    -0.39
    EDEFAULT
    -0.36
     Crock
    -0.35
    commission
    -0.35
     crock
    -0.35
    opposition
    -0.34
    UserScript
    -0.34
    don
    -0.34
     doz
    -0.33
     sesama
    -0.33
    POSITIVE LOGITS
    Segment
    0.73
     Segment
    0.69
     threshold
    0.66
    HtmlAttribute
    0.65
     hate
    0.65
     CreateTagHelper
    0.62
     seuil
    0.61
     HATE
    0.60
     Threshold
    0.60
     segment
    0.59
    Activations Density 0.074%

    No Known Activations