© 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 - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 540
    Prev
    Next
    INDEX
    Explanations

    terms related to voluntary participation and withdrawal in research studies

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

    phrases related to participant consent and withdrawal in research studies

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

    stating reason

    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
    ftagPool
    -0.58
    UserScript
    -0.55
    expandindo
    -0.54
    TestTools
    -0.50
     المعيارى
    -0.50
     estimés
    -0.48
    Filmography
    -0.47
    htra
    -0.46
     AssemblyCulture
    -0.46
    lántico
    -0.45
    POSITIVE LOGITS
     reasons
    2.86
     reason
    2.72
     Reasons
    2.55
    reason
    2.52
    reasons
    2.51
    Reason
    2.46
    Reasons
    2.43
     Reason
    2.37
     REASONS
    2.28
    REASONS
    2.13
    Activations Density 0.243%

    No Known Activations