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. 16-GEMMASCOPE-RES-65K
    6. 53014
    Prev
    Next
    INDEX
    Explanations

    programming-related syntax and keywords

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

    mathematical optimization and definitions

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
     espada
    -0.55
     texts
    -0.48
    Cab
    -0.48
    mittags
    -0.46
     planos
    -0.46
     mať
    -0.45
    side
    -0.45
     hisz
    -0.44
    cab
    -0.44
    estros
    -0.43
    POSITIVE LOGITS
    DebuggerNonUser
    6.50
    UnusedPrivate
    1.23
    DebuggerStep
    1.23
     ujednoznacz
    1.16
     ProtoMessage
    1.06
     مشين
    1.02
    WireFormatLite
    1.00
    rungsseite
    1.00
    sizeCache
    0.99
    scriptcase
    0.96
    Activations Density 0.000%

    No Known Activations