© 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. 24-GEMMASCOPE-RES-16K
    6. 2007
    Prev
    Next
    INDEX
    Explanations

    keywords related to technology and monitoring actions

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

    Tokens preceding "down"

    np_acts-logits-general · gemini-2.0-flash

    word followed by down

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Hochspringen
    -0.60
    usermodel
    -0.47
    binant
    -0.46
    Itoa
    -0.46
    orianCalendar
    -0.46
    الحياه
    -0.45
    utico
    -0.44
     ferons
    -0.44
    Digits
    -0.44
    pense
    -0.44
    POSITIVE LOGITS
     down
    4.46
    down
    3.73
    Down
    3.56
     Down
    3.47
     DOWN
    3.35
    DOWN
    3.10
    downs
    2.69
     downs
    2.50
    ダウン
    2.37
     dow
    2.14
    Activations Density 0.646%

    No Known Activations