© 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. 31-GEMMASCOPE-2-RES-16K
    7. 415
    Prev
    Next
    INDEX
    Explanations

    first time or step

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

    The neuron fires on words that mark sequence or list ordering—e.g. tokens like “first,” “next,” “one,” 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
     乀
    0.27
     Các
    0.26
     firstly
    0.25
    }
    0.25
    ）
    0.25
     Giá
    0.24
     На
    0.24
    )}{\
    0.24
     删除
    0.24
    首先
    0.24
    POSITIVE LOGITS
     iteration
    0.57
     foray
    0.54
     incarnation
    0.53
     instalment
    0.53
     installment
    0.52
     hurdle
    0.49
     iterations
    0.46
     thing
    0.44
    iteration
    0.44
     week
    0.43
    Activations Density 1.263%

    No Known Activations