© 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. 13929
    Prev
    Next
    INDEX
    Explanations

    words followed by specific completions

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

    The neuron fires on the key “topic” word or heading in formally structured or enumerated text (e.g. the main noun in a numbered clause, section heading, date label, or similar list element).

    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
    ar
    0.84
     to
    0.71
    ath
    0.69
    am
    0.68
    特殊
    0.68
    أ
    0.68
    漠
    0.67
    重要的
    0.67
    掂
    0.66
    撇
    0.65
    POSITIVE LOGITS
    oğlu
    1.03
     οι
    0.88
    ﻚ
    0.88
     τό
    0.84
    𝗎
    0.82
     डेढ़
    0.81
     그러
    0.81
     निर्मित
    0.80
     griev
    0.79
     jis
    0.79
    Activations Density 0.000%

    No Known Activations