© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 671
    Prev
    Next
    INDEX
    Explanations

    document, improving, shortening

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

    The neuron is keyed to subword fragments (prefixes or suffixes) of longer compound or technical terms—e.g. parts of “mobiler,” “parasympathomimetic,” “post‐…, ” “handels,” etc.

    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
    ၅
    0.63
    ORICAL
    0.62
     дорогие
    0.61
     धमाकेदार
    0.61
     іх
    0.61
     буквы
    0.61
    ၄
    0.60
    commutative
    0.59
    新た
    0.59
    IZATION
    0.59
    POSITIVE LOGITS
     segment
    0.53
     program
    0.52
     related
    0.49
     center
    0.48
     관련
    0.47
     kapsamında
    0.46
     पद्धतीने
    0.46
     liitty
    0.45
    0.45
     system
    0.45
    Activations Density 0.260%

    No Known Activations