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-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 5481
    Prev
    Next
    INDEX
    Explanations

    specific technical or scientific terms and identifiers

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

    biological, medical, or social contexts

    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
    AddTagHelper
    -0.88
    الحياه
    -0.76
    httphttps
    -0.76
    UnsafeEnabled
    -0.73
     CreateTagHelper
    -0.73
    SequentialGroup
    -0.72
    setVerticalGroup
    -0.72
     protoimpl
    -0.71
    AndEndTag
    -0.71
    tagHelperRunner
    -0.69
    POSITIVE LOGITS
    skrä
    0.35
    ukur
    0.32
    ījum
    0.31
    #
    0.31
    ímite
    0.30
     disediakan
    0.30
    cimentos
    0.30
    <
    0.29
     cucchiaio
    0.29
    RootElement
    0.29
    Activations Density 0.026%

    No Known Activations