© 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. 21-GEMMASCOPE-TRANSCODER-16K
    7. 2763
    Prev
    Next
    INDEX
    Explanations

    This neuron has mixed behavior, finding words related to sewing, code/computing, time/dates, reasons and requirements.

    oai_token-act-pair · gemini-2.0-flash

    programming-related terms

    np_token-act-pair-logits · gpt-4o-mini
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.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
    Due
    -0.74
     Due
    -0.73
    due
    -0.67
     DUE
    -0.61
     due
    -0.56
     inasmuch
    -0.51
    DUE
    -0.49
    由于
    -0.48
     mengingat
    -0.47
     devido
    -0.46
    POSITIVE LOGITS
    ISupport
    0.65
    存于互联网档案馆
    0.63
    ExecuteReader
    0.63
     IBOutlet
    0.61
    othed
    0.61
    anganese
    0.61
    LoadScene
    0.61
     כי
    0.60
    detectChanges
    0.59
    AddTagHelper
    0.59
    Activations Density 0.929%

    No Known Activations