© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 8113
    Prev
    Next
    INDEX
    Explanations

    deforestation &

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

    This neuron fires on technical or procedural language—tokens that mark steps, methods, or instructions (e.g. “step,” “map out,” “precedence constraint,” “data path,” “pattern,” 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_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
     mula
    0.66
    8
    0.62
     Manner
    0.62
    .‏
    0.59
    。
    0.58
    Typography
    0.57
     решила
    0.57
    6
    0.57
    logout
    0.55
    netlify
    0.55
    POSITIVE LOGITS
     或者
    0.66
     って
    0.65
    👮
    0.63
     included
    0.63
     কিংবা
    0.62
     adicionales
    0.62
    ーリー
    0.62
     इन्हीं
    0.61
    ってます
    0.61
     혹은
    0.61
    Activations Density 0.890%

    No Known Activations