© 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. 18-GEMMASCOPE-RES-16K
    6. 8820
    Prev
    Next
    INDEX
    Explanations

    phrases related to social etiquette and manners regarding doors and entrances

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

    "door"

    np_acts-logits-general · gemini-2.0-flash

    door opens/closes/locked/unlocked

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    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
    -0.63
    ]--;
    -0.58
    Билгалдахарш
    -0.48
    badah
    -0.47
    upuesto
    -0.46
     navigu
    -0.45
    CITATION
    -0.45
    tld
    -0.44
    othelioma
    -0.44
    acres
    -0.43
    POSITIVE LOGITS
     door
    1.79
     doors
    1.75
     locks
    1.54
     Door
    1.52
    door
    1.49
     lock
    1.48
     DOOR
    1.47
     Doors
    1.47
    Door
    1.47
     locking
    1.44
    Activations Density 0.217%

    No Known Activations