© 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
    2. Gemma-3-12B
    3. 31-GEMMASCOPE-2-RES-16K
    4. 6317
    Prev
    Next
    INDEX
    Explanations

    Google team and entities

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

    This neuron spots named-entity tokens—especially parts of organization or website names (e.g. Find Mugshots .com), agency labels (“law enforcement agencies”), and other proper‐noun entity words.

    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
     by
    2.69
     oleh
    2.61
     توسط
    2.28
    โดย
    2.06
    by
    2.03
    由
    1.74
     By
    1.65
     দ্বারা
    1.64
    By
    1.59
     द्वारा
    1.56
    POSITIVE LOGITS
    means
    0.66
     means
    0.66
     using
    0.62
     standards
    0.61
     selaku
    0.60
     Means
    0.58
     pursuant
    0.58
     during
    0.57
     utilizando
    0.57
     Standards
    0.56
    Activations Density 0.367%

    No Known Activations