© 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. 20-GEMMASCOPE-RES-16K
    6. 13315
    Prev
    Next
    INDEX
    Explanations

    code structures and programming syntax elements like variable declarations and function definitions

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

    Appears before or after a number

    np_acts-logits-general · gemini-2.0-flash

    code constructs and identifiers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
     
    -0.57
    …
    -0.56
     himo
    -0.55
    <em>
    -0.55
    <sup>
    -0.55
    and
    -0.54
     ra
    -0.51
    <eos>
    -0.51
    ah
    -0.50
    ,
    -0.49
    POSITIVE LOGITS
    存于互联网档案馆
    0.84
     myſelf
    0.82
    carteles
    0.77
    Betyg
    0.76
    ArgsConstructor
    0.73
    enumi
    0.72
    batore
    0.71
    Tikang
    0.70
    ulents
    0.70
    featureID
    0.69
    Activations Density 0.334%

    No Known Activations