© 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-2-2B
    4. Residual Stream - 16k
    5. 4-GEMMASCOPE-RES-16K
    6. 3
    Prev
    Next
    INDEX
    Explanations

    dates and years written in numerical form.

    oai_token-act-pair · gemini-1.5-flashTriggered by @zz11

    references to specific dates and times

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

    months of the year followed by 4 digit years.

    oai_token-act-pair · gemini-1.5-proTriggered by @zz11
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.4.hook_resid_post
    Hook Layer
    4
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    /**
    -0.65
    createNewFile
    -0.61
    ไหน
    -0.61
    الدراسه
    -0.59
     AssemblyCompany
    -0.57
    setDo
    -0.57
    zade
    -0.54
    intios
    -0.54
     Located
    -0.54
    SearchParams
    -0.54
    POSITIVE LOGITS
     ddelweddau
    0.51
    허
    0.51
     مشين
    0.50
     pinulongan
    0.50
     during
    0.49
     periods
    0.49
     nonatomic
    0.48
     late
    0.47
    شهاد
    0.46
    PreferredItem
    0.46
    Activations Density 0.667%

    No Known Activations