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-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 3794
    Prev
    Next
    INDEX
    Explanations

    structured data and object initialization in code

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

    Code involving file paths, dates, or scene kit

    np_acts-logits-general · gemini-2.0-flash

    file path construction

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    FunctionFlags
    -0.98
    FormTagHelper
    -0.80
     дописавши
    -0.77
    setVerticalGroup
    -0.75
     himo
    -0.68
     Roskov
    -0.67
     Савезне
    -0.65
    ContentAsync
    -0.64
    adpleegd
    -0.64
     mergeFrom
    -0.64
    POSITIVE LOGITS
    ⃗
    0.48
    ittaa
    0.47
     $('<
    0.45
    ches
    0.45
    NSIndexPath
    0.45
     Touren
    0.43
     CGSize
    0.43
     formado
    0.43
    inā
    0.43
    <bos>
    0.42
    Activations Density 0.337%

    No Known Activations