© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    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. 6107
    Prev
    Next
    INDEX
    Explanations

    technical terms related to persistence management in software development

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

    Code or descriptions related to implementations

    np_acts-logits-general · gemini-2.0-flash

    implementations and methods

    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
    Personendaten
    -0.79
    EndGlobalSection
    -0.75
    !*\
    -0.59
    TextSpan
    -0.57
    testng
    -0.55
    AlterField
    -0.54
    nologue
    -0.53
     errorMsg
    -0.53
    Portale
    -0.52
    GEBURTS
    -0.52
    POSITIVE LOGITS
     implemented
    0.83
     implementing
    0.76
    抽象
    0.76
     abstractions
    0.75
     implement
    0.75
     handling
    0.74
     implementations
    0.73
     implementation
    0.72
     implements
    0.72
    implemented
    0.71
    Activations Density 1.026%

    No Known Activations