© 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. 34-GEMMASCOPE-RES-131K
    7. 99276
    Prev
    Next
    INDEX
    Explanations

    describes a compound or existence

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

    chemical compounds and scientific terminology in technical documents.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron consistently lights up on numerical tokens (e.g. measurements, counts, numeric values, and chemical formula numbers).

    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
    kız
    -0.93
     Sı
    -0.93
    嘣
    -0.92
    談社
    -0.85
     dirigeants
    -0.84
    luit
    -0.82
    嚓
    -0.82
    吽
    -0.82
    били
    -0.81
     Böyle
    -0.81
    POSITIVE LOGITS
     colorless
    1.12
     readily
    1.09
     exists
    1.08
     occurs
    1.02
     white
    1.01
     stable
    0.98
     commercial
    0.97
     used
    0.96
     existence
    0.95
     forms
    0.95
    Activations Density 0.097%

    No Known Activations