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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 8640
    Prev
    Next
    INDEX
    Explanations

    specific numerical arrays and their structural characteristics

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

    numpy array manipulations

    np_acts-logits-general · gemini-2.0-flash

    numpy array operations

    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
    ,
    -0.57
     FetchType
    -0.53
    onAttach
    -0.52
    slidesToShow
    -0.51
    ContentAlignment
    -0.50
    ...
    -0.49
    OGND
    -0.48
    Carcinogenicity
    -0.46
    டுத்த
    -0.46
    MoveNext
    -0.44
    POSITIVE LOGITS
    //
    0.75
    ->__
    0.73
     חיצוניים
    0.66
     $_"
    0.65
     Xa
    0.65
    NUMX
    0.65
     ſte
    0.65
     (_.
    0.64
     KRE
    0.64
     Ibis
    0.63
    Activations Density 0.042%

    No Known Activations