© 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. 20-GEMMASCOPE-RES-131K
    6. 90639
    Prev
    Next
    INDEX
    Explanations

    inquiries about examples and ideas for various topics

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

    asking for ways or ideas

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

    phrases asking for ideas, suggestions, or ways to do something (requests for advice).

    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
     nonUne
    -0.38
     مشين
    -0.36
    jącej
    -0.35
     combineReducers
    -0.34
     Anspruch
    -0.34
    🥈
    -0.33
     ſta
    -0.33
     jaka
    -0.32
     kyse
    -0.32
     postIndex
    -0.31
    POSITIVE LOGITS
     CreateTagHelper
    0.69
     ways
    0.67
    IntoConstraints
    0.63
    Ways
    0.60
    ----</
    0.57
     Ways
    0.56
     suggestions
    0.56
     examples
    0.54
     ideas
    0.54
     WAYS
    0.51
    Activations Density 0.105%

    No Known Activations