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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 5720
    Prev
    Next
    INDEX
    Explanations

    statements regarding truths or claims and their validity

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

    Following "is" or similar words

    np_acts-logits-general · gemini-2.0-flash

    is true or important

    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-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
     Wiktionnaire
    -0.71
    InjectAttribute
    -0.69
    OGND
    -0.63
    MLLoader
    -0.60
    SharedDtor
    -0.58
    ioutil
    -0.50
    DockStyle
    -0.50
    kheim
    -0.49
    #!/
    -0.49
    ftagPool
    -0.48
    POSITIVE LOGITS
     true
    2.52
    true
    2.22
     True
    2.04
    True
    2.02
     TRUE
    1.86
    TRUE
    1.68
     vrai
    1.46
     truer
    1.33
    false
    1.27
     verdadeiro
    1.27
    Activations Density 0.386%

    No Known Activations