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. 22-GEMMASCOPE-RES-16K
    6. 15656
    Prev
    Next
    INDEX
    Explanations

    references to political behaviors, parties, and societal norms

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

    Followed by a verb

    np_acts-logits-general · gemini-2.0-flash

    to followed by a verb

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Whilst
    -1.20
     Whilst
    -1.14
    非常的
    -1.06
     whilst
    -1.01
     içerisinde
    -0.93
     poichè
    -0.89
     welches
    -0.87
    十分的
    -0.83
     utilising
    -0.82
     dimana
    -0.81
    POSITIVE LOGITS
     freilich
    0.93
     ―――――
    0.82
    essentiel
    0.76
     NDEBUG
    0.76
     guère
    0.76
    <_>
    0.75
    ditto
    0.74
     ")";
    0.73
    >*/
    0.72
    eabouts
    0.71
    Activations Density 1.146%

    No Known Activations