Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 637
    Prev
    Next
    INDEX
    Explanations

    phrases indicating emotional reactions or states of longing

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

    Punctuation followed by certain function words

    np_acts-logits-general · gemini-2.0-flash

    following comma

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Архиви
    -0.63
    apunov
    -0.58
    ligg
    -0.58
     AUSSI
    -0.57
     Cedex
    -0.56
    drawiam
    -0.55
     anteriormente
    -0.55
     précédemment
    -0.54
    等
    -0.54
     >=",
    -0.54
    POSITIVE LOGITS
    MigrationBuilder
    0.57
    Dữ
    0.57
     voices
    0.56
    SuccessListener
    0.55
     or
    0.53
     henvisninger
    0.53
     مشين
    0.52
    XtraEditors
    0.51
     zelfs
    0.50
     even
    0.50
    Activations Density 0.309%

    No Known Activations