© 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. 24-GEMMASCOPE-2-RES-16K
    7. 3178
    Prev
    Next
    INDEX
    Explanations

    Knowledge graph integration specific code

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

    The neuron strongly lights up only on rare or unusual tokens—i.e. low‐frequency or out-of-vocabulary 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_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
     Eurasian
    0.74
     BOD
    0.71
    issä
    0.71
     thắng
    0.70
     nghệ
    0.68
     Vase
    0.67
     Î
    0.67
     Musa
    0.66
     khat
    0.66
     cheats
    0.66
    POSITIVE LOGITS
    듭
    0.80
    рта
    0.77
    Prothorax
    0.77
    Transl
    0.76
     عمومی
    0.76
    Plenty
    0.74
    Generating
    0.74
    ب
    0.74
    대로
    0.73
    ORTS
    0.73
    Activations Density 0.000%

    No Known Activations