© 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. 8118
    Prev
    Next
    INDEX
    Explanations

    words related to numerical expressions and fractions

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

    Numbers spelled out with "teen", "ty", or "ths" afterward

    np_acts-logits-general · gemini-2.0-flash

    number words and fractions

    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
    3
    -0.43
    2
    -0.43
    1
    -0.42
     
    -0.38
     lata
    -0.37
     grig
    -0.36
     médica
    -0.36
     vieja
    -0.36
    leaning
    -0.36
     libre
    -0.35
    POSITIVE LOGITS
    IntoConstraints
    0.99
     الحره
    0.91
    AndEndTag
    0.85
     صوتيه
    0.85
    DockStyle
    0.84
    دانشنامهٔ
    0.80
    makeConstraints
    0.78
    ValueStyle
    0.77
    enumii
    0.75
     BorderRadius
    0.74
    Activations Density 0.130%

    No Known Activations