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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 3656
    Prev
    Next
    INDEX
    Explanations

    references to doorways or, more generally, entering/exiting

    oai_token-act-pair · gemini-2.0-flash

    doors

    np_max-act-logits · gemini-2.0-flash

    the

    np_max-act-logits · gpt-5-miniTriggered by @chenshw0109
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     door
    -1.57
    door
    -1.55
     Door
    -1.48
    Door
    -1.47
     doors
    -1.42
     DOOR
    -1.39
    doors
    -1.36
     Doors
    -1.35
     gates
    -1.34
    Doors
    -1.32
    POSITIVE LOGITS
     يتيمه
    0.49
    /***/
    0.49
    İstinadlar
    0.49
     оригіналу
    0.47
     swinging
    0.47
    Swing
    0.47
     swung
    0.47
    Erreferentziak
    0.46
    __':
    0.46
     automatiques
    0.45
    Activations Density 0.895%

    No Known Activations