© 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. 2990
    Prev
    Next
    INDEX
    Explanations

    numbers and replacement

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

    The neuron strongly activates on numeric tokens representing large negative or decimal-style scores/measurements.

    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
     भूमि
    0.61
    illery
    0.58
     agat
    0.57
    ِب
    0.56
     చేస్తుంది
    0.54
    斛
    0.54
    enges
    0.54
     suffers
    0.53
     sucede
    0.53
     сухо
    0.53
    POSITIVE LOGITS
    简化
    0.60
    <0xE5>
    0.59
    vip
    0.59
     הכ
    0.58
     Ayer
    0.57
     सात
    0.57
    ตัว
    0.57
    replaced
    0.55
     পয়
    0.55
     ordained
    0.55
    Activations Density 0.000%

    No Known Activations