© 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. 13876
    Prev
    Next
    INDEX
    Explanations

    A: as answer

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

    This neuron detects the beginning of an answer section, specifically the “A:” token that introduces an answer.

    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
     emporter
    -2.05
     organisiert
    -1.78
     nieuwe
    -1.63
    slightly
    -1.59
     refroid
    -1.57
     appliquer
    -1.57
     vôtre
    -1.57
     envoyer
    -1.56
     aktiviteter
    -1.56
     apporter
    -1.55
    POSITIVE LOGITS
     be
    1.97
     when
    1.87
     because
    1.80
    ]->
    1.70
     This
    1.55
     will
    1.53
     who
    1.49
     have
    1.49
     during
    1.48
    圃
    1.46
    Activations Density 0.065%

    No Known Activations