© 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. 18-GEMMASCOPE-RES-16K
    6. 5945
    Prev
    Next
    INDEX
    Explanations

    terms related to biological processes and environmental factors

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

    followed by "driven", "dependent", or "intensive"

    np_acts-logits-general · gemini-2.0-flash

    data driven

    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
    ſelf
    -0.77
    rungsseite
    -0.75
    Semitism
    -0.69
     externi
    -0.67
     Theſe
    -0.66
     BoxDecoration
    -0.65
     Jefus
    -0.65
    Datuak
    -0.64
    IMENTS
    -0.62
    disambiguation
    -0.61
    POSITIVE LOGITS
    diven
    0.50
    lessly
    0.49
     ddelweddau
    0.47
     driven
    0.45
    ństw
    0.45
     lät
    0.45
     AssemblyCulture
    0.44
    tually
    0.44
     sensitive
    0.43
    styleable
    0.42
    Activations Density 0.346%

    No Known Activations