© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 1999
    Prev
    Next
    INDEX
    Explanations

    The neuron is triggered by standalone uppercase letter tokens or short letter‐sequence labels (like “A”, “AB”, “F” or Roman‐numeral chords such as “V/IV”) used to tag codes or sections.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     ۲۰۰
    0.67
     paraphrase
    0.59
     puțin
    0.56
     liturg
    0.56
     ۲۰
    0.56
     sient
    0.55
    khidmat
    0.55
    📝
    0.55
    
    0.55
     நடத்த
    0.55
    POSITIVE LOGITS
    abc
    0.93
    AB
    0.87
    ABC
    0.85
    AA
    0.84
    abcd
    0.83
    abcdef
    0.82
    AC
    0.77
    BB
    0.77
    AG
    0.76
    CC
    0.74
    Activations Density 0.123%

    No Known Activations