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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 64987
    Prev
    Next
    INDEX
    Explanations

    common articles and conjunctions indicating subject or action in sentences

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

    end punctuation or common separators

    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
    AndEndTag
    -0.67
     Италијани
    -0.65
    expandindo
    -0.65
    parsedMessage
    -0.62
    rrggbb
    -0.60
     StatelessWidget
    -0.59
    protoimpl
    -0.57
     Normdatei
    -0.55
     Administrativna
    -0.53
    rungsseite
    -0.53
    POSITIVE LOGITS
     Vereine
    0.40
    kijken
    0.38
    RotationOrder
    0.34
     które
    0.33
     bowiem
    0.33
    "]/
    0.32
    ')],
    0.31
     infatti
    0.30
    '}),
    0.30
    ewel
    0.30
    Activations Density 0.036%

    No Known Activations