Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit 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 - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 1136
    Prev
    Next
    INDEX
    Explanations

    references to membranes and membrane-related concepts

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

    mentions of cellular membranes and related cellular structures

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

    membrane components and types

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
     AssemblyCulture
    -0.76
     Normdatei
    -0.71
    msgTypes
    -0.69
     mío
    -0.68
     endforeach
    -0.65
     meurt
    -0.65
    UnknownFieldSet
    -0.64
     croire
    -0.62
     Ugly
    -0.62
     Halley
    -0.61
    POSITIVE LOGITS
     membranes
    0.68
    ChildScrollView
    0.65
     Membran
    0.63
     MEM
    0.55
     ExecuteAsync
    0.54
     Walkover
    0.53
    Membrane
    0.52
    ous
    0.52
    GraphicsUnit
    0.51
     iParam
    0.51
    Activations Density 0.041%

    No Known Activations