© 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. 10-GEMMASCOPE-RES-131K
    7. 20104
    Prev
    Next
    INDEX
    Explanations

    chemical formulas, hydrogen

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

    chemical formulas and molecular structures, particularly when detecting hydrogen atoms in chemical notation.

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

    The neuron is selectively detecting numeric subscripts in chemical formulas (the atom‐count numbers in molecular formulas).

    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
     because
    -2.27
     by
    -2.11
     For
    -2.05
     from
    -1.93
     when
    -1.90
     after
    -1.84
     there
    -1.82
    jeras
    -1.79
    for
    -1.76
    puted
    -1.75
    POSITIVE LOGITS
    exemple
    1.91
    в
    1.86
    mariage
    1.82
    horloge
    1.82
    coffret
    1.80
    
    1.78
     bland
    1.74
     conseillé
    1.73
     actuellement
    1.73
     mangled
    1.71
    Activations Density 0.007%

    No Known Activations