© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 39003
    Prev
    Next
    INDEX
    Explanations

    beginning of structures or sequences

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

    The neuron activates on numeric tokens—particularly standalone numbers and measurements (e.g. digits, pixel values, circuit counts). It’s essentially a “number/measurement” detector.

    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
    ULING
    -0.93
    latter
    -0.84
     ブレスレット
    -0.81
     scelte
    -0.81
     compagni
    -0.76
    Materiales
    -0.76
     نیز
    -0.76
     ночью
    -0.74
    Ꮒ
    -0.74
     izvē
    -0.73
    POSITIVE LOGITS
     first
    1.02
     начале
    0.94
     before
    0.91
    תח
    0.89
     antes
    0.89
     davor
    0.88
    FormGroup
    0.86
     vorher
    0.86
     bran
    0.85
     prior
    0.84
    Activations Density 0.047%

    No Known Activations