© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 8512
    Prev
    Next
    INDEX
    Explanations

    I/we need

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

    The neuron fires on the first‐person singular pronoun “I,” i.e. spots where the author is speaking about themselves.

    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
     commissioned
    0.65
     periods
    0.58
     resided
    0.55
     inventive
    0.55
     possessing
    0.55
     enthusiasts
    0.54
     whilst
    0.54
    Ћ
    0.53
     enthusiast
    0.52
     inhomogeneities
    0.52
    POSITIVE LOGITS
     usually
    0.77
     সাধারণত
    0.75
     Usually
    0.73
    usually
    0.71
     biasanya
    0.71
     Обычно
    0.69
    Usually
    0.67
     Biasanya
    0.67
     تقوم
    0.66
     ALWAYS
    0.66
    Activations Density 0.282%

    No Known Activations