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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 66822
    Prev
    Next
    INDEX
    Explanations

    important principles and concepts related to integrity and support in various contexts

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

    integrity of, principle of, preferred form, childbearing intentions

    np_acts-logits-general · gemini-2.5-flash-lite

    words that signal official statements, policies, principles, goals or other formal/authoritative claims.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
    msgSender
    -0.36
    nonumber
    -0.36
    知
    -0.34
    UserScript
    -0.34
     pousser
    -0.34
     Spins
    -0.33
     ͜ʖ
    -0.33
     kick
    -0.33
    лак
    -0.32
    Allowance
    -0.31
    POSITIVE LOGITS
    :✨
    0.77
    ViewFeatures
    0.57
     Administrativna
    0.55
    あく
    0.53
    writerow
    0.52
    HomeAsUpEnabled
    0.51
     ſta
    0.50
    Rohy
    0.50
    følge
    0.49
    LookAnd
    0.48
    Activations Density 0.067%

    No Known Activations