© 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. 25815
    Prev
    Next
    INDEX
    Explanations

    numbers followed by percentages or counts

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

    The neuron never fires on any of the tokens—it appears to be effectively “dead” and doesn’t detect any pattern.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    namental
    -0.94
    ostomy
    -0.94
    ព
    -0.91
     "
    -0.88
    bottomrule
    -0.88
    szön
    -0.86
    ΑΡ
    -0.85
    waffle
    -0.83
    喩
    -0.82
     even
    -0.81
    POSITIVE LOGITS
     cucharadita
    0.94
    Kuva
    0.91
    Pourquoi
    0.91
    けれ
    0.90
    も多く
    0.89
    Déf
    0.88
    Deskripsi
    0.88
    という感じ
    0.88
    というか
    0.87
     יותר
    0.87
    Activations Density 0.012%

    No Known Activations