© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128318
    Prev
    Next
    INDEX
    Explanations

    The neuron responds to hyphen or dash tokens (i.e. the “–” or “-” characters) within the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    specialized eyeglasses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Вита
    -1.20
    ائدة
    -1.11
     (
    -1.11
     although
    -1.05
     despite
    -1.03
    asties
    -1.02
    かは
    -1.01
     extenso
    -0.99
    礎
    -0.98
     actually
    -0.97
    POSITIVE LOGITS
     takich
    1.28
     the
    1.19
    СКИ
    1.15
     tendrán
    1.13
     deberán
    1.12
     lanç
    1.11
     hefty
    1.10
     pierwszym
    1.08
    ?$
    1.08
    almo
    1.08
    Activations Density 0.113%

    No Known Activations