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. 1430
    Prev
    Next
    INDEX
    Explanations

    terms related to mathematical properties and conditions

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

    Abstract nouns followed by "of," "is," or "can."

    np_acts-logits-general · gemini-2.0-flash

    qualities, consequences, or aspects

    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
     such
    -0.66
    such
    -0.64
     thereon
    -0.61
     scolaires
    -0.55
    oredCriteria
    -0.53
     this
    -0.52
    älj
    -0.52
     tels
    -0.52
     telles
    -0.52
     astfel
    -0.51
    POSITIVE LOGITS
    AddressBook
    0.67
     resourceCulture
    0.62
    ngdoc
    0.58
     piece
    0.56
     👆
    0.56
    MenuView
    0.55
    StructEnd
    0.55
     alone
    0.54
    tula
    0.54
    fromnode
    0.54
    Activations Density 0.594%

    No Known Activations