© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 9-GEMMASCOPE-RES-131K
    6. 119789
    Prev
    Next
    INDEX
    Explanations

    patterns of various underscore-prefixed tokens in code-related contexts

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

    programming keywords and file types

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.9.hook_resid_post
    Hook Layer
    9
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Inscrivez
    -0.64
    pecabe
    -0.60
     avoient
    -0.57
     miniaturka
    -0.56
    #
    -0.54
     Suivez
    -0.53
     serce
    -0.52
     dům
    -0.52
     Komunikasi
    -0.52
     tričko
    -0.51
    POSITIVE LOGITS
     Mad
    0.58
     Savage
    0.56
    Ugly
    0.56
    ugly
    0.54
     Nice
    0.54
    Nice
    0.52
     Famous
    0.51
    bald
    0.50
    minded
    0.50
    Mad
    0.50
    Activations Density 0.087%

    No Known Activations