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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 1517
    Prev
    Next
    INDEX
    Explanations

    patterns and phrases related to questioning and inquiry

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

    Following "in", "to", "of", "via"

    np_acts-logits-general · gemini-2.0-flash

    abundance, plentiful, high concentrations

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    yonel
    -0.44
    揄
    -0.43
    itif
    -0.42
    ingdon
    -0.41
    edged
    -0.41
    isière
    -0.41
    brada
    -0.40
     Magister
    -0.40
    terscotch
    -0.40
    umet
    -0.39
    POSITIVE LOGITS
     getItemCount
    0.49
     abundant
    0.49
     abund
    0.46
     abundance
    0.46
    abund
    0.44
     abundancia
    0.44
     plentiful
    0.42
     abundante
    0.41
     frequent
    0.39
     abond
    0.39
    Activations Density 0.594%

    No Known Activations