© 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. 34-GEMMASCOPE-RES-131K
    7. 2614
    Prev
    Next
    INDEX
    Explanations

    I'm sorry

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

    The neuron fires on first-person emotional statements, i.e. “I …” expressions conveying personal sentiment.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     excuses
    -0.91
    stem
    -0.91
    salt
    -0.88
    SAND
    -0.88
     excuse
    -0.85
    saving
    -0.84
    school
    -0.83
     Tema
    -0.82
     excus
    -0.82
    shub
    -0.82
    POSITIVE LOGITS
     sorry
    2.52
     sor
    1.91
     Sor
    1.34
    Sorry
    1.22
    Sor
    1.13
     Sorry
    1.13
    SOR
    1.13
    sor
    1.05
     siento
    1.04
     sorption
    1.03
    Activations Density 0.026%

    No Known Activations