© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 10345
    Prev
    Next
    INDEX
    Explanations

    providing instructions for

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

    The neuron is tracking how far into the input sequence a token occurs (i.e. it’s effectively a learned “position” detector whose activation decays as you move through the text).

    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_24_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
     OBJECTTYPE
    0.91
     bolesti
    0.91
     LOTRAchievement
    0.90
     principalTable
    0.88
     principalColumn
    0.87
     keySchedule
    0.86
    녠
    0.85
    าที
    0.84
    <0x19>
    0.84
     glTexParameteri
    0.83
    POSITIVE LOGITS
    S
    1.09
    s
    1.08
    y
    0.95
    C
    0.92
    L
    0.91
    V
    0.91
    R
    0.88
     gon
    0.86
    H
    0.85
    Up
    0.83
    Activations Density 0.000%

    No Known Activations