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. 16-GEMMASCOPE-RES-16K
    6. 6553
    Prev
    Next
    INDEX
    Explanations

    specific names and terms related to people, places, or entities within the context of various subjects

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

    with unusual capitalization/punctuation

    np_acts-logits-general · gemini-2.0-flash

    Guggenheim, Tchetgen, Gepner

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Monfieur
    -1.11
     myſelf
    -1.08
     itſelf
    -1.05
     Theſe
    -0.96
     estekak
    -0.95
     niedersachsen
    -0.93
     ſeveral
    -0.93
     becauſe
    -0.92
     preſent
    -0.92
     ―――――
    -0.92
    POSITIVE LOGITS
    ha
    0.69
    nar
    0.63
    ak
    0.62
    ij
    0.62
    enn
    0.60
    ris
    0.59
     Nar
    0.57
    da
    0.57
     Lok
    0.56
    dos
    0.56
    Activations Density 0.945%

    No Known Activations