© 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. 26352
    Prev
    Next
    INDEX
    Explanations

    number followed by a comma

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

    The neuron lights up on specialized technical or domain‐specific nouns (e.g. jargon, field‐specific terms, and proper names).

    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
     for
    -0.85
     fået
    -0.83
    endront
    -0.82
    olução
    -0.81
     slow
    -0.79
    ].)
    -0.77
     slowed
    -0.76
     it
    -0.76
     birlikte
    -0.76
    ]
    -0.76
    POSITIVE LOGITS
    ÁT
    0.96
    viä
    0.94
    ển
    0.93
    푣
    0.88
    OBJET
    0.87
    iast
    0.87
     pratiquement
    0.87
    rennt
    0.87
     aught
    0.87
    자인
    0.86
    Activations Density 0.242%

    No Known Activations