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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 31-GEMMASCOPE-RES-131K
    6. 44709
    Prev
    Next
    INDEX
    Explanations

    specific programming constructs and data types within code

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

    data and computational contexts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     sure
    -1.13
    sure
    -1.02
    Sure
    -0.91
     ensure
    -0.87
     Sure
    -0.86
     ensured
    -0.86
     SURE
    -0.85
     confirm
    -0.79
     certeza
    -0.78
     ensures
    -0.78
    POSITIVE LOGITS
     publicités
    0.43
    ftagPool
    0.37
    createComponent
    0.36
    getJSONArray
    0.35
    Datuak
    0.34
    __":
    0.33
    цезда
    0.33
    dicionado
    0.33
    eqn
    0.32
    Disclosure
    0.31
    Activations Density 0.059%

    No Known Activations