© 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. 10-GEMMASCOPE-RES-131K
    7. 67262
    Prev
    Next
    INDEX
    Explanations

    typo and typographical errors

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron flags any word containing the substring “typo” (e.g. typo, typographical, typological, typologies).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     is
    -3.23
     were
    -2.63
     The
    -2.61
    s
    -2.42
    ll
    -2.42
    ↵
    -2.34
    StoryboardSegue
    -2.19
     provides
    -2.08
    which
    -2.06
    we
    -2.05
    POSITIVE LOGITS
     problemet
    2.41
    geſ
    2.34
    gelijk
    2.19
    皷
    2.13
     truthful
    2.11
    鷽
    2.08
    0
    2.05
    räder
    2.02
     círculos
    2.00
     elektrik
    1.98
    Activations Density 0.004%

    No Known Activations